File size: 1,223 Bytes
f79fc82
8bce4c2
 
 
 
 
 
 
 
 
 
f79fc82
496e059
18969d2
 
 
 
 
 
3a600a3
18969d2
3a600a3
18969d2
3a600a3
18969d2
3a600a3
18969d2
3a600a3
18969d2
3a600a3
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

---
title: Synthetic Data Gen
emoji: πŸ‘
colorFrom: indigo
colorTo: yellow
sdk: gradio
sdk_version: 5.9.0
app_file: app.py
pinned: false
short_description: Synthetic data generation for training using diffusion model
---

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference


Features:

Stable Diffusion Integration: Leverages the StableDiffusionImg2ImgPipeline for versatile image generation.

Class Misclassification Analysis: Automatically identifies misclassified class pairs for focused image generation.

Advanced Prompt Interpolation: Utilizes spherical linear interpolation (slerp) to generate smooth transitions between text embeddings for more natural image outputs.

Structural Similarity Scoring: Calculates SSIM scores to measure the quality of generated images compared to input references.

Comprehensive Metadata Management: Captures extensive details about the generation process for easy reproducibility and debugging.

Dynamic Directory Management: Automatically creates and organizes directories for class pairs and metadata files.

Highly Configurable: Supports advanced optimization options like torch.compile and DeepCache for performance improvements.