DR-TANet: Dynamic Receptive Temporal Attention Network for Street Scene Change Detection

TL;DR

This paper introduces DR-TANet, a novel network for street scene change detection that employs dynamic receptive temporal attention and concurrent horizontal and vertical attention to improve accuracy and efficiency.

Contribution

It proposes a new temporal attention module and a lightweight version, DRTAM, along with CHVA, enhancing change detection performance and efficiency in street scenes.

Findings

Achieves state-of-the-art results on GSV, TSUNAMI, and VL-CMU-CD datasets.

Maintains high efficiency suitable for autonomous vehicle applications.

Outperforms existing methods without complex enhancements.

Abstract

Street scene change detection continues to capture researchers' interests in the computer vision community. It aims to identify the changed regions of the paired street-view images captured at different times. The state-of-the-art network based on the encoder-decoder architecture leverages the feature maps at the corresponding level between two channels to gain sufficient information of changes. Still, the efficiency of feature extraction, feature correlation calculation, even the whole network requires further improvement. This paper proposes the temporal attention and explores the impact of the dependency-scope size of temporal attention on the performance of change detection. In addition, based on the Temporal Attention Module (TAM), we introduce a more efficient and light-weight version - Dynamic Receptive Temporal Attention Module (DRTAM) and propose the Concurrent Horizontal and Vertical Attention (CHVA) to improve the accuracy of the network on specific challenging entities. On street scene datasets `GSV', `TSUNAMI' and `VL-CMU-CD', our approach gains excellent performance, establishing new state-of-the-art scores without bells and whistles, while maintaining high efficiency applicable in autonomous vehicles.