C-DIRA: Computationally Efficient Dynamic ROI Routing and Domain-Invariant Adversarial Learning for Lightweight Driver Behavior Recognition

TL;DR

C-DIRA is a novel lightweight framework for driver behavior recognition that combines dynamic ROI routing and domain-invariant adversarial learning to improve accuracy, efficiency, and robustness across diverse conditions.

Contribution

It introduces a dynamic ROI routing mechanism and domain-invariant adversarial training to enhance lightweight driver behavior recognition models.

Findings

Achieves high accuracy with fewer FLOPs and lower latency.

Demonstrates robustness to visual degradation like blur and low-light.

Maintains stable performance across unseen domains.

Abstract

Driver distraction behavior recognition using in-vehicle cameras demands real-time inference on edge devices. However, lightweight models often fail to capture fine-grained behavioral cues, resulting in reduced performance on unseen drivers or under varying conditions. ROI-based methods also increase computational cost, making it difficult to balance efficiency and accuracy. This work addresses the need for a lightweight architecture that overcomes these constraints. We propose Computationally efficient Dynamic region of Interest Routing and domain-invariant Adversarial learning for lightweight driver behavior recognition (C-DIRA). The framework combines saliency-driven Top-K ROI pooling and fused classification for local feature extraction and integration. Dynamic ROI routing enables selective computation by applying ROI inference only to high difficulty data samples. Moreover, pseudo-domain labeling and adversarial learning are used to learn domain-invariant features robust to driver and background variation. Experiments on the State Farm Distracted Driver Detection Dataset show that C-DIRA maintains high accuracy with significantly fewer FLOPs and lower latency than prior lightweight models. It also demonstrates robustness under visual degradation such as blur and low-light, and stable performance across unseen domains. These results confirm C-DIRA's effectiveness in achieving compactness, efficiency, and generalization.