2nd Place Solution for Waymo Open Dataset Challenge -- 2D Object Detection

TL;DR

This paper presents a robust 2D object detection system for autonomous driving that combines multiple models and ensemble techniques, achieving high accuracy on the Waymo dataset and securing second place in the challenge.

Contribution

It introduces a hybrid detection framework integrating two-stage and one-stage anchor-free detectors with an auto ensemble scheme for improved accuracy.

Findings

Achieved 70.28 L2 mAP on Waymo Open Dataset v1.2

Secured 2nd place in the Waymo Open Dataset Challenge

Demonstrated effectiveness of model ensemble in autonomous driving detection

Abstract

A practical autonomous driving system urges the need to reliably and accurately detect vehicles and persons. In this report, we introduce a state-of-the-art 2D object detection system for autonomous driving scenarios. Specifically, we integrate both popular two-stage detector and one-stage detector with anchor free fashion to yield a robust detection. Furthermore, we train multiple expert models and design a greedy version of the auto ensemble scheme that automatically merges detections from different models. Notably, our overall detection system achieves 70.28 L2 mAP on the Waymo Open Dataset v1.2, ranking the 2nd place in the 2D detection track of the Waymo Open Dataset Challenges.