🚗 VideoMAE-2 for Dashcam Collision Prediction

This repository contains model for predicting vehicle collisions using dashcam footage, developed for the Nexar Dashcam Collision Prediction Challenge. The model achieved a 11th place finish on the leaderboard with a score of 0.80 mean Average Precision (mAP)

For Training code - GitHub

---

🧠 Model Overview

• Architecture: VideoMAE-2 Large fine-tuned for binary classification (collision/near-miss vs. normal driving).

• Feature Extraction: Utilized TimeSformer for preprocessing input frames.

• Input: 16 frames per video, each resized to 224x224 pixels.

• Output: Probability score indicating the likelihood of a collision or near-miss event.

---

📁 Dataset

The model was trained on the Nexar Collision Prediction Dataset.

• 750 non-collision videos

• 400 collision videos

• 350 near-miss videos arXiv

Each video is annotated with:

• Event Type: Collision, near-miss, or normal driving

• Event Time: Timestamp of the (near-)collision

• Alert Time: Earliest time the event could be predicted.

For more details, refer to the dataset paper.

---

🛠️ Preprocessing Pipeline

1. Frame Extraction: Sampled 16 frames per video, focusing on the interval around the alert time.

2. Feature Extraction: Applied TimeSformer feature extractor to obtain pixel values.

3. Data Augmentation: Implemented transformations such as horizontal flip, rotation, color jitter, and resized cropping.

4. Normalization: Used ImageNet mean and standard deviation for normalization.

---

🏋️ Training Details

• Framework: PyTorch with Hugging Face Transformers and Trainer API.

• Training Configuration:

  • Batch Size: 4

  • Epochs: 15

  • Learning Rate: 3e-5

  • Weight Decay: 0.01

  • Evaluation Strategy: Per epoch

  • Metric for Best Model: Average Precision

• Hardware: Trained on 2x NVIDIA T4 GPUs (~4.5 hours)

---

📊 Evaluation Metrics

The model's performance was evaluated using Mean Average Precision (mAP) across different time-to-accident intervals:

• 500ms

• 1000ms

• 1500ms

The final score is the mean of the Average Precision (AP) values at these intervals, emphasizing early and accurate collision predictions

📚 Citation

If you use this model or dataset, please cite:

@misc{nexar2025dashcamcollisionprediction,
  title={Nexar Dashcam Collision Prediction Dataset and Challenge},
  author={Daniel C. Moura and Shizhan Zhu and Orly Zvitia},
  year={2025},
  eprint={2503.03848},
  archivePrefix={arXiv},
  primaryClass={cs.CV},
  url={https://arxiv.org/abs/2503.03848}
}