Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,7 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+output.jpg filter=lfs diff=lfs merge=lfs -text
+output_augmentation.jpg filter=lfs diff=lfs merge=lfs -text
+output_confusion_matrix.png filter=lfs diff=lfs merge=lfs -text
+results.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,97 @@
+---
+license: mit
+library_name: ultralytics
+tags:
+- image-classification
+- yolo
+- ultralytics
+- drowsiness-detection
+- computer-vision
+widget:
+- modelId: mosesb/drowsiness-detection-yolo-cls
+  title: Drowsiness Detection With YOLO CLS
+  url: https://huggingface.co/spaces/mosesb/drowsiness-detection-yolo-cls/resolve/main/output.jpg
+datasets:
+- ismailnasri20/driver-drowsiness-dataset-ddd
+- yasharjebraeily/drowsy-detection-dataset
+metrics:
+- accuracy
+- f1
+---
+
+# YOLOv11 Model for Drowsiness Detection
+
+This repository contains a YOLO classification model fine-tuned to detect driver drowsiness from images. The model classifies input images into two categories: `Drowsy` and `Non Drowsy` (Awake).
+
+This model was trained using the `ultralytics` framework and demonstrates high performance on an unseen test set, making it a reliable tool for safety applications.
+
+## Model Details
+*   **Base Model:** `yolo11x-cls` (from the Ultralytics v8 ecosystem)
+*   **Fine-tuned on:** A combined dataset for driver drowsiness detection.
+*   **Classes:** `Drowsy`, `Non Drowsy`
+*   **Framework:** PyTorch, Ultralytics
+
+## How to Get Started
+
+You can easily use this model with the `ultralytics` library.
+
+```python
+# Install ultralytics
+!pip install ultralytics
+
+from ultralytics import YOLO
+
+# Load the model from the Hugging Face Hub
+model = YOLO('your-username/your-repo-name')
+
+# Run inference on an image
+image_path = 'path/to/your/image.jpg'
+results = model.predict(image_path)
+
+# Print the top prediction
+probs = results[0].probs
+top1_class_index = probs.top1
+top1_confidence = probs.top1conf
+class_name = model.names[top1_class_index]
+
+print(f"Prediction: {class_name} with confidence {top1_confidence:.4f}")
+```
+
+## Training Procedure
+
+The model was fine-tuned on a large dataset of driver images. The training process involved:
+-   **Data Augmentation:** Standard augmentations like random flips, color jitter (HSV), and scaling were applied.
+-   **Transfer Learning:** The model was initialized with weights pretrained on a large-scale dataset, enabling rapid convergence.
+
+### Key Hyperparameters
+- **Image Size:** 224x224
+- **Batch Size:** 185 (auto-tuned)
+- **Optimizer:** SGD with momentum
+
+![Training Results](results.png)
+
+## Evaluation
+
+The model was evaluated on a completely **unseen test set** to ensure a fair assessment of its generalization capabilities.
+
+### Key Performance Metrics
+| Metric | Value  | Description                                        |
+| :----: | :----: | :------------------------------------------------- |
+| **Accuracy** | 99.80% | Overall correctness on the test set.           |
+| **APCER**    | 0.00%  | Rate of 'Drowsy' drivers missed (False Negatives). |
+| **BPCER**    | 0.41%  | Rate of 'Non Drowsy' drivers flagged (False Positives). |
+| **ACER**     | 0.21%  | Average of APCER and BPCER.                        |
+
+*APCER (Attack Presentation Classification Error Rate) is the most critical safety metric.*
+
+![Confusion Matrix](output_confusion_matrix.png)
+
+### Model Explainability (Grad-CAM)
+To ensure the model is focusing on relevant facial features, Grad-CAM was used. The heatmaps confirm that the model's predictions are primarily based on the eye and mouth regions, as expected.
+
+![Grad-CAM](output_grad_cam.jpg)
+
+## Intended Use and Limitations
+This model is intended as a proof-of-concept for driver safety systems. It should not be used as the sole mechanism for preventing accidents. Real-world performance may vary based on lighting conditions, camera angles, occlusions (e.g., sunglasses), and individual differences.
+
+*This model card is based on the training notebook [`yolov11_drowsiness.ipynb`](https://github.com/mosesab/YOLOV11-Drowsiness-Detection/blob/main/yolov11_drowsiness.ipynb).*

app.py

@@ -0,0 +1,49 @@
+import gradio as gr
+from ultralytics import YOLO
+import torch
+
+# Load the fine-tuned YOLOv11x model
+# The model will be in the same directory in the HF Space
+model = YOLO('best.pt')
+
+# Define the prediction function
+def predict_drowsiness(image):
+    """
+    Takes a PIL image, runs inference, and returns a dictionary of class probabilities.
+    """
+    # Run prediction
+    results = model.predict(image, verbose=False)
+    
+    # Get the class names from the model
+    names_dict = results[0].names
+    
+    # Get the probabilities
+    probs = results[0].probs.data.cpu().numpy()
+    
+    # Create a dictionary of {class_name: probability}
+    return {names_dict[i]: prob for i, prob in enumerate(probs)}
+
+# --- Gradio Interface ---
+# Define the title and description for the demo
+title = "YOLOv11 Drowsiness Detection"
+description = """
+This demo showcases a fine-tuned YOLO classification model for detecting driver drowsiness.
+Upload an image of a driver, and the model will predict whether the person is 'Drowsy' or 'Non Drowsy' (Awake).
+This model was trained as detailed in the notebook below and achieves high accuracy on the test set.
+Training Notebook Repo: https://github.com/mosesab/YOLOV11-Drowsiness-Detection/blob/main/yolov11_drowsiness.ipynb
+"""
+article = "Driver fatigue is a major cause of accidents. This model analyzes facial images to predict the likelihood of drowsiness in real time."
+
+# Create the Gradio interface
+iface = gr.Interface(
+    fn=predict_drowsiness,
+    inputs=gr.Image(type="pil", label="Upload Driver Image"),
+    outputs=gr.Label(num_top_classes=2, label="Prediction"),
+    title=title,
+    description=description,
+    article=article,
+    examples=[ "sample_1.jpg", "sample_2.jpg" ]
+)
+
+# Launch the app
+iface.launch()

args.yaml

@@ -0,0 +1,105 @@
+task: classify
+mode: train
+model: yolo11x-cls.pt
+data: /workspace/dataset
+epochs: 300
+time: null
+patience: 20
+batch: -1
+imgsz: 224
+save: true
+save_period: -1
+cache: true
+device: null
+workers: 8
+project: drowsiness_training
+name: yolo_cls_run
+exist_ok: true
+pretrained: true
+optimizer: auto
+verbose: true
+seed: 0
+deterministic: true
+single_cls: false
+rect: false
+cos_lr: false
+close_mosaic: 10
+resume: false
+amp: true
+fraction: 1.0
+profile: false
+freeze: null
+multi_scale: false
+overlap_mask: true
+mask_ratio: 4
+dropout: 0.0
+val: true
+split: val
+save_json: false
+conf: null
+iou: 0.7
+max_det: 300
+half: false
+dnn: false
+plots: true
+source: null
+vid_stride: 1
+stream_buffer: false
+visualize: false
+augment: false
+agnostic_nms: false
+classes: null
+retina_masks: false
+embed: null
+show: false
+save_frames: false
+save_txt: false
+save_conf: false
+save_crop: false
+show_labels: true
+show_conf: true
+show_boxes: true
+line_width: null
+format: torchscript
+keras: false
+optimize: false
+int8: false
+dynamic: false
+simplify: true
+opset: null
+workspace: null
+nms: false
+lr0: 0.01
+lrf: 0.01
+momentum: 0.937
+weight_decay: 0.0005
+warmup_epochs: 3.0
+warmup_momentum: 0.8
+warmup_bias_lr: 0.1
+box: 7.5
+cls: 0.5
+dfl: 1.5
+pose: 12.0
+kobj: 1.0
+nbs: 64
+hsv_h: 0.015
+hsv_s: 0.7
+hsv_v: 0.4
+degrees: 0.0
+translate: 0.1
+scale: 0.5
+shear: 0.0
+perspective: 0.0
+flipud: 0.0
+fliplr: 0.5
+bgr: 0.0
+mosaic: 1.0
+mixup: 0.0
+cutmix: 0.0
+copy_paste: 0.0
+copy_paste_mode: flip
+auto_augment: randaugment
+erasing: 0.4
+cfg: null
+tracker: botsort.yaml
+save_dir: drowsiness_training/yolo_cls_run

best.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4fed29ef7a18c011b440ce470554b2eb6c54fbb7c417b7ebf8245896f1a45032
+size 57006897

last.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f844b288a66917e1e12d9ae7773a8f3dfcf168ee1c2daf458b816a47240a3fab
+size 57008241

results.csv

@@ -0,0 +1,31 @@
+epoch,time,train/loss,metrics/accuracy_top1,metrics/accuracy_top5,val/loss,lr/pg0,lr/pg1,lr/pg2
+1,36.4885,0.14788,0.98671,1,0.03707,0.00332807,0.00332807,0.00332807
+1,78.2575,0.4195,0.97034,1,0.09415,0.00331811,0.00331811,0.00331811
+2,154.248,0.07064,0.99314,1,0.02335,0.0066295,0.0066295,0.0066295
+3,229.881,0.02835,0.99678,1,0.00745,0.00991888,0.00991888,0.00991888
+4,305.269,0.01964,0.99846,1,0.00526,0.009901,0.009901,0.009901
+5,380.811,0.01352,0.99874,1,0.00512,0.009868,0.009868,0.009868
+6,456.478,0.01245,0.99832,1,0.00642,0.009835,0.009835,0.009835
+7,531.746,0.00972,0.9993,1,0.00297,0.009802,0.009802,0.009802
+8,607.184,0.01014,0.99888,1,0.00312,0.009769,0.009769,0.009769
+9,682.386,0.00689,0.99958,1,0.00191,0.009736,0.009736,0.009736
+10,757.863,0.00809,0.9986,1,0.00412,0.009703,0.009703,0.009703
+11,833.026,0.00768,0.99846,1,0.00442,0.00967,0.00967,0.00967
+12,908.255,0.00767,0.99888,1,0.00294,0.009637,0.009637,0.009637
+13,983.448,0.00702,0.9979,1,0.00534,0.009604,0.009604,0.009604
+14,1058.57,0.0091,0.99818,1,0.00499,0.009571,0.009571,0.009571
+15,1133.77,0.00801,0.99902,1,0.00208,0.009538,0.009538,0.009538
+16,1208.97,0.00812,0.99622,1,0.01176,0.009505,0.009505,0.009505
+17,1284.23,0.01047,0.99818,1,0.0056,0.009472,0.009472,0.009472
+18,1359.47,0.01251,0.9986,1,0.0035,0.009439,0.009439,0.009439
+19,1434.67,0.01087,0.99874,1,0.00518,0.009406,0.009406,0.009406
+20,1509.67,0.01463,0.99482,1,0.01421,0.009373,0.009373,0.009373
+21,1584.82,0.01296,0.99818,1,0.00519,0.00934,0.00934,0.00934
+22,1659.91,0.0123,0.99902,1,0.0033,0.009307,0.009307,0.009307
+23,1735.13,0.0137,0.99804,1,0.00708,0.009274,0.009274,0.009274
+24,1810.33,0.01558,0.9986,1,0.00429,0.009241,0.009241,0.009241
+25,1885.4,0.0152,0.9986,1,0.00395,0.009208,0.009208,0.009208
+26,1961.98,0.02328,0.99944,1,0.00325,0.009175,0.009175,0.009175
+27,2040.69,0.01656,0.99902,1,0.00282,0.009142,0.009142,0.009142
+28,2115.7,0.01849,0.99888,1,0.00395,0.009109,0.009109,0.009109
+29,2190.86,0.02094,0.99804,1,0.00647,0.009076,0.009076,0.009076