Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -13,7 +13,7 @@ dataset_info:
|
|
13 |
num_bytes: 15634112.4
|
14 |
num_examples: 4
|
15 |
download_size: 81521051
|
16 |
-
dataset_size: 81514720
|
17 |
configs:
|
18 |
- config_name: default
|
19 |
data_files:
|
@@ -21,4 +21,82 @@ configs:
|
|
21 |
path: data/train-*
|
22 |
- split: test
|
23 |
path: data/test-*
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
24 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
13 |
num_bytes: 15634112.4
|
14 |
num_examples: 4
|
15 |
download_size: 81521051
|
16 |
+
dataset_size: 81514720
|
17 |
configs:
|
18 |
- config_name: default
|
19 |
data_files:
|
|
|
21 |
path: data/train-*
|
22 |
- split: test
|
23 |
path: data/test-*
|
24 |
+
license: apache-2.0
|
25 |
+
task_categories:
|
26 |
+
- object-detection
|
27 |
+
language:
|
28 |
+
- en
|
29 |
+
tags:
|
30 |
+
- objectdetection d
|
31 |
+
- detection
|
32 |
+
- syntheticdata
|
33 |
+
- yolov8
|
34 |
+
- yolo
|
35 |
+
- labels
|
36 |
+
- labeled
|
37 |
+
- label
|
38 |
+
- indoor
|
39 |
+
- cpg
|
40 |
+
- can
|
41 |
+
size_categories:
|
42 |
+
- 1K<n<10K
|
43 |
---
|
44 |
+
Soup Can Object Detection Dataset Sample
|
45 |
+
Duality.ai just released a 1000 image dataset used to train a YOLOv8 model for object detection -- and it's 100% free!
|
46 |
+
Just [create an EDU account here](link).
|
47 |
+
This HuggingFace dataset is a 20 image and label sample, but you can get the rest at no cost by [creating a FalconCloud account](link). Once you verify your email, the link will redirect you to the dataset page.
|
48 |
+
|
49 |
+
What makes this dataset unique, useful, and capable of bridging the Sim2Real gap?
|
50 |
+
- The digital twins are not generated by AI, but instead crafted by 3D artists to be INDISTINGUISHABLE to the model from the physical-world objects. This allows the training from this data to transfer into real-world applicability
|
51 |
+
- The simulation software, called FalconEditor, can easily create thousands of images with varying lighting, posing, occlusions, backgrounds, camera positions, and more. This enables robust model training.
|
52 |
+
- The labels are created along with the data. This not only saves large amounts of time, but also ensures the labels are incredibly accurate and reliable.
|
53 |
+
|
54 |
+
|
55 |
+
Dataset Overview
|
56 |
+
This dataset consists of high-quality images of soup cans captured in various poses and lighting conditions .This dataset is structured to train and test object detection models, specifically YOLO-based and other object detection frameworks.
|
57 |
+
Why Use This Dataset?
|
58 |
+
Single Object Detection: Specifically curated for detecting soup cans, making it ideal for fine-tuning models for retail, inventory management, or robotics applications.
|
59 |
+
|
60 |
+
Varied Environments: The dataset contains images with different lighting conditions, poses, and occlusions to help solve traditional recall problems in real world object detection.
|
61 |
+
|
62 |
+
Accurate Annotations: Bounding box annotations are precise and automatically labeled in YOLO format as the data is created.
|
63 |
+
Create your own specialized data!
|
64 |
+
You can create a dataset like this but with your own digital twin! [Create an account and follow this tutorial to learn how](link).
|
65 |
+
Dataset Structure
|
66 |
+
The dataset is organized as follows:
|
67 |
+
Soup-Can-Object-Detection-Dataset/
|
68 |
+
β-- images/
|
69 |
+
β βββ 000000000.png
|
70 |
+
β βββ 000000001.png
|
71 |
+
β βββ ...
|
72 |
+
β-- labels/
|
73 |
+
β βββ 000000000.txt
|
74 |
+
β βββ 000000001.txt
|
75 |
+
β βββ ...
|
76 |
+
|
77 |
+
Components
|
78 |
+
Images: RGB images of the soup can in .png format.
|
79 |
+
Labels: .txt files containing bounding box annotations in the YOLO format.
|
80 |
+
0 = soup can
|
81 |
+
Example Annotation (YOLO Format):
|
82 |
+
|
83 |
+
0 0.475 0.554 0.050 0.050
|
84 |
+
|
85 |
+
Where:
|
86 |
+
0 represents the object class (soup can).
|
87 |
+
The next four values represent the bounding box coordinates (normalized x_center, y_center, width, height).
|
88 |
+
Usage
|
89 |
+
This dataset is designed to be used with popular deep learning frameworks:
|
90 |
+
|
91 |
+
from datasets import load_dataset
|
92 |
+
|
93 |
+
dataset = load_dataset("your-huggingface-username/Soup-Can-Object-Detection")
|
94 |
+
|
95 |
+
To train a YOLOv8 model, you can use Ultralytics' yolo package:
|
96 |
+
|
97 |
+
yolo train model=yolov8n.pt data=soup_can.yaml epochs=50 imgsz=640
|
98 |
+
|
99 |
+
|
100 |
+
Licensing
|
101 |
+
License: Apache 2.0
|
102 |
+
Attribution: If you use this dataset in research or commercial projects, please provide appropriate credit.
|