Update README.md
Browse files
README.md
CHANGED
|
@@ -11,4 +11,73 @@ license: apache-2.0
|
|
| 11 |
short_description: Comprehensive Image Evaluation Tool
|
| 12 |
---
|
| 13 |
|
| 14 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
short_description: Comprehensive Image Evaluation Tool
|
| 12 |
---
|
| 13 |
|
| 14 |
+
# Comprehensive Image Evaluation Tool
|
| 15 |
+
|
| 16 |
+
This tool combines multiple image evaluation models into a single application with a user-friendly interface for analyzing and reviewing images.
|
| 17 |
+
|
| 18 |
+
## Features
|
| 19 |
+
|
| 20 |
+
- **Batch Processing**: Upload multiple images at once for efficient evaluation
|
| 21 |
+
- **Multiple Models**: Combines evaluations from several aesthetic prediction models:
|
| 22 |
+
- ShadowLilac's aesthetic-shadow-v2
|
| 23 |
+
- WaifuScorer V4
|
| 24 |
+
- CafeAI's aesthetic, style and waifu classifiers
|
| 25 |
+
- Anime Aesthetic predictor
|
| 26 |
+
- **Comprehensive Analysis**: Get detailed metrics for each image
|
| 27 |
+
- **Results Table**: View results sorted by score with image previews
|
| 28 |
+
- **Export**: Save results to CSV for further analysis
|
| 29 |
+
- **Single Image Mode**: Evaluate individual images and get detailed results
|
| 30 |
+
|
| 31 |
+
## Installation
|
| 32 |
+
|
| 33 |
+
1. Clone this repository:
|
| 34 |
+
```
|
| 35 |
+
git clone [repository-url]
|
| 36 |
+
cd image-evaluation-tool
|
| 37 |
+
```
|
| 38 |
+
|
| 39 |
+
2. Install required dependencies:
|
| 40 |
+
```
|
| 41 |
+
pip install -r requirements.txt
|
| 42 |
+
```
|
| 43 |
+
|
| 44 |
+
3. Run the application:
|
| 45 |
+
```
|
| 46 |
+
python app.py
|
| 47 |
+
```
|
| 48 |
+
|
| 49 |
+
## Usage
|
| 50 |
+
|
| 51 |
+
### Batch Processing
|
| 52 |
+
|
| 53 |
+
1. Launch the application
|
| 54 |
+
2. Use the file upload panel to select multiple images
|
| 55 |
+
3. Adjust the HQ threshold if needed (default 0.5)
|
| 56 |
+
4. Click "Process Images"
|
| 57 |
+
5. View results in the table sorted by average score
|
| 58 |
+
6. Click "Export Results to CSV" to save the data
|
| 59 |
+
|
| 60 |
+
### Single Image Evaluation
|
| 61 |
+
|
| 62 |
+
1. Scroll down to the Single Image Evaluation section
|
| 63 |
+
2. Upload an image
|
| 64 |
+
3. Click "Evaluate"
|
| 65 |
+
4. View detailed metrics and style information
|
| 66 |
+
|
| 67 |
+
## Models Information
|
| 68 |
+
|
| 69 |
+
- **ShadowLilac** (0-1): General aesthetic quality assessment
|
| 70 |
+
- **WaifuScorer** (0-10): Specialized for anime-style images
|
| 71 |
+
- **CafeAI** (0-1): Style classification and aesthetic assessment
|
| 72 |
+
- **Anime Aesthetic** (0-10): Specialized for anime/manga art
|
| 73 |
+
|
| 74 |
+
## Output Folders
|
| 75 |
+
|
| 76 |
+
- `output/hq_folder`: Images that meet or exceed the threshold
|
| 77 |
+
- `output/lq_folder`: Images that score below the threshold
|
| 78 |
+
|
| 79 |
+
## Requirements
|
| 80 |
+
|
| 81 |
+
- Python 3.8+
|
| 82 |
+
- CUDA-compatible GPU recommended for faster processing
|
| 83 |
+
- ~4GB of disk space for model downloads (first run)
|