Cyclic Refiner: Object-Aware Temporal Representation Learning for Multi-View 3D Detection and Tracking

TL;DR

This paper introduces Cyclic Refiner, a novel object-aware temporal learning framework that enhances multi-view 3D detection and tracking by refining features through cyclic backward propagation, reducing distractor influence, and improving robustness.

Contribution

It proposes a cyclic learning mechanism for robust multi-view feature refinement and an object-aware association strategy for improved 3D detection and tracking.

Findings

Achieves consistent performance improvements on nuScenes dataset.

Enhances robustness of temporal fusion against distractors and background clutter.

Improves object awareness and tracklet association accuracy.

Abstract

We propose a unified object-aware temporal learning framework for multi-view 3D detection and tracking tasks. Having observed that the efficacy of the temporal fusion strategy in recent multi-view perception methods may be weakened by distractors and background clutters in historical frames, we propose a cyclic learning mechanism to improve the robustness of multi-view representation learning. The essence is constructing a backward bridge to propagate information from model predictions (e.g., object locations and sizes) to image and BEV features, which forms a circle with regular inference. After backward refinement, the responses of target-irrelevant regions in historical frames would be suppressed, decreasing the risk of polluting future frames and improving the object awareness ability of temporal fusion. We further tailor an object-aware association strategy for tracking based on the cyclic learning model. The cyclic learning model not only provides refined features, but also delivers finer clues (e.g., scale level) for tracklet association. The proposed cycle learning method and association module together contribute a novel and unified multi-task framework. Experiments on nuScenes show that the proposed model achieves consistent performance gains over baselines of different designs (i.e., dense query-based BEVFormer, sparse query-based SparseBEV and LSS-based BEVDet4D) on both detection and tracking evaluation.