Optimizing Anchor-based Detectors for Autonomous Driving Scenes
This work addresses the problem of efficient and accurate object detection for autonomous driving systems, representing an incremental improvement over existing methods.
The paper tackles improving anchor-based object detectors for autonomous driving scenes by adapting them for small objects in crowds and optimizing speed-accuracy trade-offs, achieving state-of-the-art results with 76.9% AP/L1 on the Waymo Open Dataset under real-time constraints.
This paper summarizes model improvements and inference-time optimizations for the popular anchor-based detectors in the scenes of autonomous driving. Based on the high-performing RCNN-RS and RetinaNet-RS detection frameworks designed for common detection scenes, we study a set of framework improvements to adapt the detectors to better detect small objects in crowd scenes. Then, we propose a model scaling strategy by scaling input resolution and model size to achieve a better speed-accuracy trade-off curve. We evaluate our family of models on the real-time 2D detection track of the Waymo Open Dataset (WOD). Within the 70 ms/frame latency constraint on a V100 GPU, our largest Cascade RCNN-RS model achieves 76.9% AP/L1 and 70.1% AP/L2, attaining the new state-of-the-art on WOD real-time 2D detection. Our fastest RetinaNet-RS model achieves 6.3 ms/frame while maintaining a reasonable detection precision at 50.7% AP/L1 and 42.9% AP/L2.