Temporal-Channel Transformer for 3D Lidar-Based Video Object Detection in Autonomous Driving

TL;DR

This paper introduces a novel Temporal-Channel Transformer that effectively models spatial, temporal, and channel relationships in Lidar data for 3D video object detection, achieving state-of-the-art results in autonomous driving scenarios.

Contribution

It proposes a new transformer architecture with separate encoder and decoder for encoding temporal-channel and spatial-channel information, enhancing 3D object detection from Lidar sequences.

Findings

Achieves state-of-the-art performance on nuScenes benchmark.

Effectively models temporal, spatial, and channel relationships.

Improves detection accuracy over existing methods.

Abstract

The strong demand of autonomous driving in the industry has lead to strong interest in 3D object detection and resulted in many excellent 3D object detection algorithms. However, the vast majority of algorithms only model single-frame data, ignoring the temporal information of the sequence of data. In this work, we propose a new transformer, called Temporal-Channel Transformer, to model the spatial-temporal domain and channel domain relationships for video object detecting from Lidar data. As a special design of this transformer, the information encoded in the encoder is different from that in the decoder, i.e. the encoder encodes temporal-channel information of multiple frames while the decoder decodes the spatial-channel information for the current frame in a voxel-wise manner. Specifically, the temporal-channel encoder of the transformer is designed to encode the information of different channels and frames by utilizing the correlation among features from different channels and frames. On the other hand, the spatial decoder of the transformer will decode the information for each location of the current frame. Before conducting the object detection with detection head, the gate mechanism is deployed for re-calibrating the features of current frame, which filters out the object irrelevant information by repetitively refine the representation of target frame along with the up-sampling process. Experimental results show that we achieve the state-of-the-art performance in grid voxel-based 3D object detection on the nuScenes benchmark.