A Comprehensive Study on Traffic Incident Detection Using Hybrid Models
The authors propose a hybrid model combining Generative Adversarial Networks (GANs) and Transformer models to address challenges in traffic incident detection, focusing on dataset imbalance and scarcity. By leveraging the strengths of both models, they aim to enhance detection performance significantly.