|
|
--- |
|
|
title: Controlnet Depth Generation |
|
|
emoji: 🌖 |
|
|
colorFrom: red |
|
|
colorTo: purple |
|
|
sdk: gradio |
|
|
sdk_version: 5.38.0 |
|
|
app_file: app.py |
|
|
pinned: false |
|
|
license: mit |
|
|
short_description: Interior design using controlnet depth model |
|
|
--- |
|
|
|
|
|
Stable Diffusion ControlNet Depth Demo |
|
|
This Space demonstrates a Stable Diffusion model combined with a ControlNet model fine-tuned for depth, and includes automatic depth map estimation from your input image. |
|
|
|
|
|
How to use: |
|
|
|
|
|
Upload an Input Image: Provide any photo (e.g., of a room, an object, a scene). The app will automatically estimate its depth map. |
|
|
|
|
|
Enter a Text Prompt: Describe the image you want to generate. The model will try to apply your prompt while respecting the structure derived from the depth map. |
|
|
|
|
|
Adjust Parameters: Experiment with "Inference Steps" and "Guidance Scale" for different results. |
|
|
|
|
|
Click "Submit" to generate the image. |
|
|
|
|
|
Model Details: |
|
|
|
|
|
Base Diffusion Model: runwayml/stable-diffusion-v1-5 (downloaded from Hugging Face Hub) |
|
|
|
|
|
ControlNet Model: Fine-tuned for depth (uploaded as ./Output_ControlNet_Finetune) |
|
|
|
|
|
Depth Estimator: Intel/dpt-hybrid-midas (downloaded from Hugging Face Hub) |
|
|
|
|
|
Note: This model is quite large, so the first generation after a "cold start" (when the Space wakes up) might take a few minutes to load the models. Subsequent generations will be faster. |
|
|
|
|
|
Enjoy! |
|
|
|
|
|
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference |
|
|
|
