2nd Place Solution for Waymo Open Dataset Challenge -- Real-time 2D Object Detection
This addresses the need for fast and accurate detection of vehicles, pedestrians, and cyclists in autonomous systems, but it is incremental as it builds on existing methods.
The paper tackled real-time 2D object detection for autonomous driving by aggregating one-stage detectors and optimizing inference with TensorRT, achieving 75.00% L1 mAP and 69.72% L2 mAP with a latency of 45.8ms/frame on a V100 GPU.
In an autonomous driving system, it is essential to recognize vehicles, pedestrians and cyclists from images. Besides the high accuracy of the prediction, the requirement of real-time running brings new challenges for convolutional network models. In this report, we introduce a real-time method to detect the 2D objects from images. We aggregate several popular one-stage object detectors and train the models of variety input strategies independently, to yield better performance for accurate multi-scale detection of each category, especially for small objects. For model acceleration, we leverage TensorRT to optimize the inference time of our detection pipeline. As shown in the leaderboard, our proposed detection framework ranks the 2nd place with 75.00% L1 mAP and 69.72% L2 mAP in the real-time 2D detection track of the Waymo Open Dataset Challenges, while our framework achieves the latency of 45.8ms/frame on an Nvidia Tesla V100 GPU.