Update model card with V2 results and V1/V2 comparison

README.md

@@ -29,12 +29,22 @@ Binary classifier: detects whether a document image contains moiré artifacts
 - `clean` (0): No moiré patterns
 - `moire` (1): Moiré patterns detected
 
-## V2 Improvements (over V1)
-- **Larger model:** DeiT-small (22M params) vs DeiT-tiny (5.5M)
-- **More training data:** 8,000 samples vs 6,000
-- **6 moiré methods** (added subtle + localized patterns as hard examples)
-- **Label smoothing** (0.05) for better calibration
-- **Stronger augmentation** (rotation, hue jitter)
+## V1 → V2 Comparison
+
+| | V1 (DeiT-tiny) | **V2 (DeiT-small)** |
+|---|---|---|
+| **Parameters** | 5.5M | **22M** |
+| **Training samples** | 6,000 | **8,000** |
+| **Moiré methods** | 4 | **6** (+subtle, +localized) |
+| **Label smoothing** | — | **0.05** |
+| **Accuracy** | 99.5% | **99.1%** |
+| **F1 Score** | 0.995 | **0.991** |
+| **Precision** | 99.3% | **98.5%** |
+| **Recall** | 99.7% | **99.8%** |
+
+> **Note:** V2 was evaluated on harder examples including subtle single-frequency moiré and localized
+> moiré patterns that V1 never trained on. V2 achieves near-perfect recall (99.75%) — it catches
+> virtually all moiré patterns including very subtle ones, at the cost of slightly lower precision.
 
 ## Training Details
 
@@ -43,11 +53,13 @@ Binary classifier: detects whether a document image contains moiré artifacts
 | Base model | `facebook/deit-small-patch16-224` |
 | Parameters | 22M |
 | Training samples | 8,000 (4,000 clean + 4,000 moiré) |
-| Test samples | 800 (400 clean + 400 moiré) |
+| Eval samples | 800 (400 clean + 400 moiré) |
 | Epochs | 5 |
-| Learning rate | 3e-05 (cosine schedule) |
+| Learning rate | 3e-5 (cosine schedule) |
 | Effective batch size | 64 |
 | Label smoothing | 0.05 |
+| Warmup steps | 60 |
+| Best checkpoint | Epoch 2 (by F1) |
 
 ### Moiré Generation Methods
 1. **Resize aliasing** — downscale+upscale with NEAREST interpolation + pattern overlay
@@ -57,14 +69,25 @@ Binary classifier: detects whether a document image contains moiré artifacts
 5. **Subtle moiré** — very low strength single-frequency (hard examples)
 6. **Localized moiré** — moiré in elliptical region with gaussian mask
 
-## Performance
+## Performance (Best Checkpoint)
 
-| Metric | Eval |
-|--------|------|
-| Accuracy | 0.9912 |
+| Metric | Value |
+|--------|-------|
+| Accuracy | 99.12% |
 | F1 Score | 0.9913 |
-| Precision | 0.9852 |
-| Recall | 0.9975 |
+| Precision | 98.52% |
+| Recall | 99.75% |
+| Eval Loss | 0.170 |
+
+### Training Progression
+
+| Epoch | Eval Loss | Accuracy | F1 | Precision | Recall |
+|-------|-----------|----------|-----|-----------|--------|
+| 1 | 0.173 | 99.0% | 0.990 | 99.3% | 98.8% |
+| 2 | 0.177 | 99.1% | **0.991** | 99.0% | 99.3% |
+| 3 | 0.154 | 99.1% | 0.991 | 99.3% | 99.0% |
+| 4 | 0.170 | 99.1% | **0.991** | 98.5% | **99.8%** |
+| 5 | 0.168 | 99.0% | 0.990 | 98.3% | **99.8%** |
 
 ## Usage
 
@@ -90,11 +113,13 @@ image = Image.open("document.jpg")
 inputs = processor(image, return_tensors="pt")
 with torch.no_grad():
     logits = model(**inputs).logits
-    label = model.config.id2label[logits.argmax(-1).item()]
-print(label)  # 'clean' or 'moire'
+    predicted = logits.argmax(-1).item()
+
+print(model.config.id2label[predicted])  # 'clean' or 'moire'
 ```
 
 ## Limitations
-- Trained on synthetic moiré — may not capture all real-world variations
-- Optimized for document images; natural scenes may vary
-- Input resized to 224×224; very subtle moiré in high-res images may be lost
+- Trained on synthetic moiré patterns — may not capture all real-world moiré variations
+- Optimized for document images; performance on natural scene images may vary
+- Input images resized to 224×224; very subtle moiré in high-resolution images may be lost
+- Higher recall than precision — may occasionally flag clean images as moiré (false positive rate ~1.5%)