Datasets:

CodeGoat24
/

VideoDPO

Dataset card Data Studio Files Files and versions Community

Split (1)

train · 10k rows

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.

prompt stringlengths 4 8.37k	conversations listlengths 2 2	images sequencelengths 16 16
create a video where the flash and batman fight 2 minutes long.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [create a video where the flash and batman fight 2 minutes long.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/006675/0.png", "winvideos/006675/1.png", "winvideos/006675/2.png", "winvideos/006675/3.png", "winvideos/006675/4.png", "winvideos/006675/5.png", "winvideos/006675/6.png", "winvideos/006675/7.png", "losevideos/006675/0.png", "losevideos/006675/1.png", "losevideos/006675/2.png", "losevideos/006675/3.png", "losevideos/006675/4.png", "losevideos/006675/5.png", "losevideos/006675/6.png", "losevideos/006675/7.png" ]
garden of Babylon.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [garden of Babylon.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/000603/0.png", "winvideos/000603/1.png", "winvideos/000603/2.png", "winvideos/000603/3.png", "winvideos/000603/4.png", "winvideos/000603/5.png", "winvideos/000603/6.png", "winvideos/000603/7.png", "losevideos/000603/0.png", "losevideos/000603/1.png", "losevideos/000603/2.png", "losevideos/000603/3.png", "losevideos/000603/4.png", "losevideos/000603/5.png", "losevideos/000603/6.png", "losevideos/000603/7.png" ]
happy mouse pixar style with red ballon.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [happy mouse pixar style with red ballon.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/005051/0.png", "winvideos/005051/1.png", "winvideos/005051/2.png", "winvideos/005051/3.png", "winvideos/005051/4.png", "winvideos/005051/5.png", "winvideos/005051/6.png", "winvideos/005051/7.png", "losevideos/005051/0.png", "losevideos/005051/1.png", "losevideos/005051/2.png", "losevideos/005051/3.png", "losevideos/005051/4.png", "losevideos/005051/5.png", "losevideos/005051/6.png", "losevideos/005051/7.png" ]
Wide close-up, centered, daytime, an alien with a lizard head, lizard skin, walking standing still on a street of residences, unreal engine -neg “deform, distort, figure, add character, add, very fast movements, talk, talking, strange movements, beams of light, rays of light, change of scenery, rear view, side camera, rotation, no movement, unrealistic movement, add hair, moving food, moving objects, rear view, side camera, rotation, no movement, unreal movement” -motion 1 -gs 12 -fps 24 -hide.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [Wide close-up, centered, daytime, an alien with a lizard head, lizard skin, walking standing still on a street of residences, unreal engine -neg “deform, distort, figure, add character, add, very fast movements, talk, talking, strange movements, beams of light, rays of light, change of scenery, rear view, side camera, rotation, no movement, unrealistic movement, add hair, moving food, moving objects, rear view, side camera, rotation, no movement, unreal movement” -motion 1 -gs 12 -fps 24 -hide.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/001391/0.png", "winvideos/001391/1.png", "winvideos/001391/2.png", "winvideos/001391/3.png", "winvideos/001391/4.png", "winvideos/001391/5.png", "winvideos/001391/6.png", "winvideos/001391/7.png", "losevideos/001391/0.png", "losevideos/001391/1.png", "losevideos/001391/2.png", "losevideos/001391/3.png", "losevideos/001391/4.png", "losevideos/001391/5.png", "losevideos/001391/6.png", "losevideos/001391/7.png" ]
the ninja swings his sword, and from this come Japanese candles in the background against the backdrop of a bear and a bull fighting /video.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [the ninja swings his sword, and from this come Japanese candles in the background against the backdrop of a bear and a bull fighting /video.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/008069/0.png", "losevideos/008069/1.png", "losevideos/008069/2.png", "losevideos/008069/3.png", "losevideos/008069/4.png", "losevideos/008069/5.png", "losevideos/008069/6.png", "losevideos/008069/7.png", "winvideos/008069/0.png", "winvideos/008069/1.png", "winvideos/008069/2.png", "winvideos/008069/3.png", "winvideos/008069/4.png", "winvideos/008069/5.png", "winvideos/008069/6.png", "winvideos/008069/7.png" ]
hands, marriage.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [hands, marriage.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/000949/0.png", "losevideos/000949/1.png", "losevideos/000949/2.png", "losevideos/000949/3.png", "losevideos/000949/4.png", "losevideos/000949/5.png", "losevideos/000949/6.png", "losevideos/000949/7.png", "winvideos/000949/0.png", "winvideos/000949/1.png", "winvideos/000949/2.png", "winvideos/000949/3.png", "winvideos/000949/4.png", "winvideos/000949/5.png", "winvideos/000949/6.png", "winvideos/000949/7.png" ]
manbird talking to satan.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [manbird talking to satan.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/007597/0.png", "winvideos/007597/1.png", "winvideos/007597/2.png", "winvideos/007597/3.png", "winvideos/007597/4.png", "winvideos/007597/5.png", "winvideos/007597/6.png", "winvideos/007597/7.png", "losevideos/007597/0.png", "losevideos/007597/1.png", "losevideos/007597/2.png", "losevideos/007597/3.png", "losevideos/007597/4.png", "losevideos/007597/5.png", "losevideos/007597/6.png", "losevideos/007597/7.png" ]
Portrait of a magipunk city, moving parts, anime. By greg rutkowski, greg tocchini, james gilleard, joe fenton, kaethe butcher, gradient pink, cyan, white, teal color scheme, grunge aesthetic!!! trending on artstation, caravaggio, greg rutkowski, beeple, beksinski, giger, pascal blanche rutkowski.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [Portrait of a magipunk city, moving parts, anime. By greg rutkowski, greg tocchini, james gilleard, joe fenton, kaethe butcher, gradient pink, cyan, white, teal color scheme, grunge aesthetic!!! trending on artstation, caravaggio, greg rutkowski, beeple, beksinski, giger, pascal blanche rutkowski.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/008187/0.png", "winvideos/008187/1.png", "winvideos/008187/2.png", "winvideos/008187/3.png", "winvideos/008187/4.png", "winvideos/008187/5.png", "winvideos/008187/6.png", "winvideos/008187/7.png", "losevideos/008187/0.png", "losevideos/008187/1.png", "losevideos/008187/2.png", "losevideos/008187/3.png", "losevideos/008187/4.png", "losevideos/008187/5.png", "losevideos/008187/6.png", "losevideos/008187/7.png" ]
a ninja in a cinematic fight scene.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [a ninja in a cinematic fight scene.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/006112/0.png", "winvideos/006112/1.png", "winvideos/006112/2.png", "winvideos/006112/3.png", "winvideos/006112/4.png", "winvideos/006112/5.png", "winvideos/006112/6.png", "winvideos/006112/7.png", "losevideos/006112/0.png", "losevideos/006112/1.png", "losevideos/006112/2.png", "losevideos/006112/3.png", "losevideos/006112/4.png", "losevideos/006112/5.png", "losevideos/006112/6.png", "losevideos/006112/7.png" ]
falling petals of roses falling cinematic lighting.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [falling petals of roses falling cinematic lighting.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/000811/0.png", "winvideos/000811/1.png", "winvideos/000811/2.png", "winvideos/000811/3.png", "winvideos/000811/4.png", "winvideos/000811/5.png", "winvideos/000811/6.png", "winvideos/000811/7.png", "losevideos/000811/0.png", "losevideos/000811/1.png", "losevideos/000811/2.png", "losevideos/000811/3.png", "losevideos/000811/4.png", "losevideos/000811/5.png", "losevideos/000811/6.png", "losevideos/000811/7.png" ]
lion bouncing on a trampoline.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [lion bouncing on a trampoline.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/005573/0.png", "winvideos/005573/1.png", "winvideos/005573/2.png", "winvideos/005573/3.png", "winvideos/005573/4.png", "winvideos/005573/5.png", "winvideos/005573/6.png", "winvideos/005573/7.png", "losevideos/005573/0.png", "losevideos/005573/1.png", "losevideos/005573/2.png", "losevideos/005573/3.png", "losevideos/005573/4.png", "losevideos/005573/5.png", "losevideos/005573/6.png", "losevideos/005573/7.png" ]
I'm a mystical object, a treasure untold. Find me in stories, ancient and old. Wizards wield me with powers untamed, in the hands of heroes, I'm often named. What am I?.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [I'm a mystical object, a treasure untold. Find me in stories, ancient and old. Wizards wield me with powers untamed, in the hands of heroes, I'm often named. What am I?.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/008802/0.png", "losevideos/008802/1.png", "losevideos/008802/2.png", "losevideos/008802/3.png", "losevideos/008802/4.png", "losevideos/008802/5.png", "losevideos/008802/6.png", "losevideos/008802/7.png", "winvideos/008802/0.png", "winvideos/008802/1.png", "winvideos/008802/2.png", "winvideos/008802/3.png", "winvideos/008802/4.png", "winvideos/008802/5.png", "winvideos/008802/6.png", "winvideos/008802/7.png" ]
lego rap video.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [lego rap video.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/001233/0.png", "losevideos/001233/1.png", "losevideos/001233/2.png", "losevideos/001233/3.png", "losevideos/001233/4.png", "losevideos/001233/5.png", "losevideos/001233/6.png", "losevideos/001233/7.png", "winvideos/001233/0.png", "winvideos/001233/1.png", "winvideos/001233/2.png", "winvideos/001233/3.png", "winvideos/001233/4.png", "winvideos/001233/5.png", "winvideos/001233/6.png", "winvideos/001233/7.png" ]
With a surge of brilliance, seven fairies overcome the witch's malevolence.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [With a surge of brilliance, seven fairies overcome the witch's malevolence.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/008873/0.png", "losevideos/008873/1.png", "losevideos/008873/2.png", "losevideos/008873/3.png", "losevideos/008873/4.png", "losevideos/008873/5.png", "losevideos/008873/6.png", "losevideos/008873/7.png", "winvideos/008873/0.png", "winvideos/008873/1.png", "winvideos/008873/2.png", "winvideos/008873/3.png", "winvideos/008873/4.png", "winvideos/008873/5.png", "winvideos/008873/6.png", "winvideos/008873/7.png" ]
$100 bills are raining down from the sky.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [$100 bills are raining down from the sky.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/004108/0.png", "losevideos/004108/1.png", "losevideos/004108/2.png", "losevideos/004108/3.png", "losevideos/004108/4.png", "losevideos/004108/5.png", "losevideos/004108/6.png", "losevideos/004108/7.png", "winvideos/004108/0.png", "winvideos/004108/1.png", "winvideos/004108/2.png", "winvideos/004108/3.png", "winvideos/004108/4.png", "winvideos/004108/5.png", "winvideos/004108/6.png", "winvideos/004108/7.png" ]
futuristic enemies fighting on a vast battlefield.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [futuristic enemies fighting on a vast battlefield.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/005913/0.png", "winvideos/005913/1.png", "winvideos/005913/2.png", "winvideos/005913/3.png", "winvideos/005913/4.png", "winvideos/005913/5.png", "winvideos/005913/6.png", "winvideos/005913/7.png", "losevideos/005913/0.png", "losevideos/005913/1.png", "losevideos/005913/2.png", "losevideos/005913/3.png", "losevideos/005913/4.png", "losevideos/005913/5.png", "losevideos/005913/6.png", "losevideos/005913/7.png" ]
animated childrens book, come to life.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [animated childrens book, come to life.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/009466/0.png", "losevideos/009466/1.png", "losevideos/009466/2.png", "losevideos/009466/3.png", "losevideos/009466/4.png", "losevideos/009466/5.png", "losevideos/009466/6.png", "losevideos/009466/7.png", "winvideos/009466/0.png", "winvideos/009466/1.png", "winvideos/009466/2.png", "winvideos/009466/3.png", "winvideos/009466/4.png", "winvideos/009466/5.png", "winvideos/009466/6.png", "winvideos/009466/7.png" ]
prompt:workmates teasing a colleague in the work -seed 8746046.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [prompt:workmates teasing a colleague in the work -seed 8746046.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/004514/0.png", "winvideos/004514/1.png", "winvideos/004514/2.png", "winvideos/004514/3.png", "winvideos/004514/4.png", "winvideos/004514/5.png", "winvideos/004514/6.png", "winvideos/004514/7.png", "losevideos/004514/0.png", "losevideos/004514/1.png", "losevideos/004514/2.png", "losevideos/004514/3.png", "losevideos/004514/4.png", "losevideos/004514/5.png", "losevideos/004514/6.png", "losevideos/004514/7.png" ]
cinematic shot of a Greek warrior reading the letter.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [cinematic shot of a Greek warrior reading the letter.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/001131/0.png", "losevideos/001131/1.png", "losevideos/001131/2.png", "losevideos/001131/3.png", "losevideos/001131/4.png", "losevideos/001131/5.png", "losevideos/001131/6.png", "losevideos/001131/7.png", "winvideos/001131/0.png", "winvideos/001131/1.png", "winvideos/001131/2.png", "winvideos/001131/3.png", "winvideos/001131/4.png", "winvideos/001131/5.png", "winvideos/001131/6.png", "winvideos/001131/7.png" ]
The Gioconda drinking a glass of wine, Monna Lisa drinks wine.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [The Gioconda drinking a glass of wine, Monna Lisa drinks wine.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/006574/0.png", "winvideos/006574/1.png", "winvideos/006574/2.png", "winvideos/006574/3.png", "winvideos/006574/4.png", "winvideos/006574/5.png", "winvideos/006574/6.png", "winvideos/006574/7.png", "losevideos/006574/0.png", "losevideos/006574/1.png", "losevideos/006574/2.png", "losevideos/006574/3.png", "losevideos/006574/4.png", "losevideos/006574/5.png", "losevideos/006574/6.png", "losevideos/006574/7.png" ]
a vintage hair horn is blowing and a red light is flashing and turning.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [a vintage hair horn is blowing and a red light is flashing and turning.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/008203/0.png", "losevideos/008203/1.png", "losevideos/008203/2.png", "losevideos/008203/3.png", "losevideos/008203/4.png", "losevideos/008203/5.png", "losevideos/008203/6.png", "losevideos/008203/7.png", "winvideos/008203/0.png", "winvideos/008203/1.png", "winvideos/008203/2.png", "winvideos/008203/3.png", "winvideos/008203/4.png", "winvideos/008203/5.png", "winvideos/008203/6.png", "winvideos/008203/7.png" ]
background is swirling gas and flashing colors: a pentagram appears in the West: make it one pentagram.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [background is swirling gas and flashing colors: a pentagram appears in the West: make it one pentagram.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/003794/0.png", "losevideos/003794/1.png", "losevideos/003794/2.png", "losevideos/003794/3.png", "losevideos/003794/4.png", "losevideos/003794/5.png", "losevideos/003794/6.png", "losevideos/003794/7.png", "winvideos/003794/0.png", "winvideos/003794/1.png", "winvideos/003794/2.png", "winvideos/003794/3.png", "winvideos/003794/4.png", "winvideos/003794/5.png", "winvideos/003794/6.png", "winvideos/003794/7.png" ]
a pretty girl in a photostudio, full body shot, colorful and bright, -ar 9:16.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [a pretty girl in a photostudio, full body shot, colorful and bright, -ar 9:16.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/007805/0.png", "winvideos/007805/1.png", "winvideos/007805/2.png", "winvideos/007805/3.png", "winvideos/007805/4.png", "winvideos/007805/5.png", "winvideos/007805/6.png", "winvideos/007805/7.png", "losevideos/007805/0.png", "losevideos/007805/1.png", "losevideos/007805/2.png", "losevideos/007805/3.png", "losevideos/007805/4.png", "losevideos/007805/5.png", "losevideos/007805/6.png", "losevideos/007805/7.png" ]
pizza wave.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [pizza wave.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/004993/0.png", "winvideos/004993/1.png", "winvideos/004993/2.png", "winvideos/004993/3.png", "winvideos/004993/4.png", "winvideos/004993/5.png", "winvideos/004993/6.png", "winvideos/004993/7.png", "losevideos/004993/0.png", "losevideos/004993/1.png", "losevideos/004993/2.png", "losevideos/004993/3.png", "losevideos/004993/4.png", "losevideos/004993/5.png", "losevideos/004993/6.png", "losevideos/004993/7.png" ]
Fire stones falling fro the sky.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [Fire stones falling fro the sky.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/009145/0.png", "losevideos/009145/1.png", "losevideos/009145/2.png", "losevideos/009145/3.png", "losevideos/009145/4.png", "losevideos/009145/5.png", "losevideos/009145/6.png", "losevideos/009145/7.png", "winvideos/009145/0.png", "winvideos/009145/1.png", "winvideos/009145/2.png", "winvideos/009145/3.png", "winvideos/009145/4.png", "winvideos/009145/5.png", "winvideos/009145/6.png", "winvideos/009145/7.png" ]
a spongebob animation.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [a spongebob animation.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/009850/0.png", "winvideos/009850/1.png", "winvideos/009850/2.png", "winvideos/009850/3.png", "winvideos/009850/4.png", "winvideos/009850/5.png", "winvideos/009850/6.png", "winvideos/009850/7.png", "losevideos/009850/0.png", "losevideos/009850/1.png", "losevideos/009850/2.png", "losevideos/009850/3.png", "losevideos/009850/4.png", "losevideos/009850/5.png", "losevideos/009850/6.png", "losevideos/009850/7.png" ]
flowers Message: moshy (Font: MODERN).	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [flowers Message: moshy (Font: MODERN).]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/002631/0.png", "losevideos/002631/1.png", "losevideos/002631/2.png", "losevideos/002631/3.png", "losevideos/002631/4.png", "losevideos/002631/5.png", "losevideos/002631/6.png", "losevideos/002631/7.png", "winvideos/002631/0.png", "winvideos/002631/1.png", "winvideos/002631/2.png", "winvideos/002631/3.png", "winvideos/002631/4.png", "winvideos/002631/5.png", "winvideos/002631/6.png", "winvideos/002631/7.png" ]
((Girls jumping on the bed, soft bed, central composition, bright, bedroom, warm colors, wide-angle lens, high-speed shutter, relaxed, and free.)) -ar 1:1.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [((Girls jumping on the bed, soft bed, central composition, bright, bedroom, warm colors, wide-angle lens, high-speed shutter, relaxed, and free.)) -ar 1:1.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/002725/0.png", "losevideos/002725/1.png", "losevideos/002725/2.png", "losevideos/002725/3.png", "losevideos/002725/4.png", "losevideos/002725/5.png", "losevideos/002725/6.png", "losevideos/002725/7.png", "winvideos/002725/0.png", "winvideos/002725/1.png", "winvideos/002725/2.png", "winvideos/002725/3.png", "winvideos/002725/4.png", "winvideos/002725/5.png", "winvideos/002725/6.png", "winvideos/002725/7.png" ]
an attractive news opening with the words "we love you".	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [an attractive news opening with the words "we love you".]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/009072/0.png", "winvideos/009072/1.png", "winvideos/009072/2.png", "winvideos/009072/3.png", "winvideos/009072/4.png", "winvideos/009072/5.png", "winvideos/009072/6.png", "winvideos/009072/7.png", "losevideos/009072/0.png", "losevideos/009072/1.png", "losevideos/009072/2.png", "losevideos/009072/3.png", "losevideos/009072/4.png", "losevideos/009072/5.png", "losevideos/009072/6.png", "losevideos/009072/7.png" ]
cinematic style looking through window to a forest, raining.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [cinematic style looking through window to a forest, raining.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/007959/0.png", "losevideos/007959/1.png", "losevideos/007959/2.png", "losevideos/007959/3.png", "losevideos/007959/4.png", "losevideos/007959/5.png", "losevideos/007959/6.png", "losevideos/007959/7.png", "winvideos/007959/0.png", "winvideos/007959/1.png", "winvideos/007959/2.png", "winvideos/007959/3.png", "winvideos/007959/4.png", "winvideos/007959/5.png", "winvideos/007959/6.png", "winvideos/007959/7.png" ]
animals in the forest at night caught on trail cam, realistic with flash.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [animals in the forest at night caught on trail cam, realistic with flash.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/001377/0.png", "winvideos/001377/1.png", "winvideos/001377/2.png", "winvideos/001377/3.png", "winvideos/001377/4.png", "winvideos/001377/5.png", "winvideos/001377/6.png", "winvideos/001377/7.png", "losevideos/001377/0.png", "losevideos/001377/1.png", "losevideos/001377/2.png", "losevideos/001377/3.png", "losevideos/001377/4.png", "losevideos/001377/5.png", "losevideos/001377/6.png", "losevideos/001377/7.png" ]
Knitting balls Message: SUHIH (Font: MODERN).	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [Knitting balls Message: SUHIH (Font: MODERN).]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/006758/0.png", "losevideos/006758/1.png", "losevideos/006758/2.png", "losevideos/006758/3.png", "losevideos/006758/4.png", "losevideos/006758/5.png", "losevideos/006758/6.png", "losevideos/006758/7.png", "winvideos/006758/0.png", "winvideos/006758/1.png", "winvideos/006758/2.png", "winvideos/006758/3.png", "winvideos/006758/4.png", "winvideos/006758/5.png", "winvideos/006758/6.png", "winvideos/006758/7.png" ]
timelaps, aerial view over the roofs of the city of innsbruck while sunrise, cinematic, anime, comic book illustration style.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [timelaps, aerial view over the roofs of the city of innsbruck while sunrise, cinematic, anime, comic book illustration style.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/008720/0.png", "losevideos/008720/1.png", "losevideos/008720/2.png", "losevideos/008720/3.png", "losevideos/008720/4.png", "losevideos/008720/5.png", "losevideos/008720/6.png", "losevideos/008720/7.png", "winvideos/008720/0.png", "winvideos/008720/1.png", "winvideos/008720/2.png", "winvideos/008720/3.png", "winvideos/008720/4.png", "winvideos/008720/5.png", "winvideos/008720/6.png", "winvideos/008720/7.png" ]
abs.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [abs.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/003506/0.png", "winvideos/003506/1.png", "winvideos/003506/2.png", "winvideos/003506/3.png", "winvideos/003506/4.png", "winvideos/003506/5.png", "winvideos/003506/6.png", "winvideos/003506/7.png", "losevideos/003506/0.png", "losevideos/003506/1.png", "losevideos/003506/2.png", "losevideos/003506/3.png", "losevideos/003506/4.png", "losevideos/003506/5.png", "losevideos/003506/6.png", "losevideos/003506/7.png" ]
James Harden dunks on Luka Donci.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [James Harden dunks on Luka Donci.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/001115/0.png", "losevideos/001115/1.png", "losevideos/001115/2.png", "losevideos/001115/3.png", "losevideos/001115/4.png", "losevideos/001115/5.png", "losevideos/001115/6.png", "losevideos/001115/7.png", "winvideos/001115/0.png", "winvideos/001115/1.png", "winvideos/001115/2.png", "winvideos/001115/3.png", "winvideos/001115/4.png", "winvideos/001115/5.png", "winvideos/001115/6.png", "winvideos/001115/7.png" ]
Its skin was a sickly pale, almost translucent, and its features were twisted into a grotesque mask of hatred and malice, Its hair, long and tangled, hung in matted strands around its face, and its fingers were thin and bony, like the claws of some demonic creature, As it floated towards me, its ethereal form seemed to shift and writhe, as if it were made of smoke and shadows, the very air around it grew cold and still, and I felt a sense of dread and foreboding wash over me like a wave, this was no mere apparition, but a malevolent force from beyond the veil, a harbinger of doom and despair. --ar 9:16.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [Its skin was a sickly pale, almost translucent, and its features were twisted into a grotesque mask of hatred and malice, Its hair, long and tangled, hung in matted strands around its face, and its fingers were thin and bony, like the claws of some demonic creature, As it floated towards me, its ethereal form seemed to shift and writhe, as if it were made of smoke and shadows, the very air around it grew cold and still, and I felt a sense of dread and foreboding wash over me like a wave, this was no mere apparition, but a malevolent force from beyond the veil, a harbinger of doom and despair. --ar 9:16.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/006500/0.png", "winvideos/006500/1.png", "winvideos/006500/2.png", "winvideos/006500/3.png", "winvideos/006500/4.png", "winvideos/006500/5.png", "winvideos/006500/6.png", "winvideos/006500/7.png", "losevideos/006500/0.png", "losevideos/006500/1.png", "losevideos/006500/2.png", "losevideos/006500/3.png", "losevideos/006500/4.png", "losevideos/006500/5.png", "losevideos/006500/6.png", "losevideos/006500/7.png" ]
5 women doing skydiving.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [5 women doing skydiving.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/004919/0.png", "losevideos/004919/1.png", "losevideos/004919/2.png", "losevideos/004919/3.png", "losevideos/004919/4.png", "losevideos/004919/5.png", "losevideos/004919/6.png", "losevideos/004919/7.png", "winvideos/004919/0.png", "winvideos/004919/1.png", "winvideos/004919/2.png", "winvideos/004919/3.png", "winvideos/004919/4.png", "winvideos/004919/5.png", "winvideos/004919/6.png", "winvideos/004919/7.png" ]
goku is playing ball.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [goku is playing ball.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/006322/0.png", "losevideos/006322/1.png", "losevideos/006322/2.png", "losevideos/006322/3.png", "losevideos/006322/4.png", "losevideos/006322/5.png", "losevideos/006322/6.png", "losevideos/006322/7.png", "winvideos/006322/0.png", "winvideos/006322/1.png", "winvideos/006322/2.png", "winvideos/006322/3.png", "winvideos/006322/4.png", "winvideos/006322/5.png", "winvideos/006322/6.png", "winvideos/006322/7.png" ]
a 1920 footage about an all coven of auvergnates women with traditional cult outfit dancing in the forest with mouth open.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [a 1920 footage about an all coven of auvergnates women with traditional cult outfit dancing in the forest with mouth open.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/005718/0.png", "losevideos/005718/1.png", "losevideos/005718/2.png", "losevideos/005718/3.png", "losevideos/005718/4.png", "losevideos/005718/5.png", "losevideos/005718/6.png", "losevideos/005718/7.png", "winvideos/005718/0.png", "winvideos/005718/1.png", "winvideos/005718/2.png", "winvideos/005718/3.png", "winvideos/005718/4.png", "winvideos/005718/5.png", "winvideos/005718/6.png", "winvideos/005718/7.png" ]
How a grown man with a skull in his hand walks through town and enters a toy store.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [How a grown man with a skull in his hand walks through town and enters a toy store.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/006903/0.png", "losevideos/006903/1.png", "losevideos/006903/2.png", "losevideos/006903/3.png", "losevideos/006903/4.png", "losevideos/006903/5.png", "losevideos/006903/6.png", "losevideos/006903/7.png", "winvideos/006903/0.png", "winvideos/006903/1.png", "winvideos/006903/2.png", "winvideos/006903/3.png", "winvideos/006903/4.png", "winvideos/006903/5.png", "winvideos/006903/6.png", "winvideos/006903/7.png" ]