STRNet: Visual Navigation with Spatio-Temporal Representation through Dynamic Graph Aggregation

TL;DR

This paper introduces STRNet, a novel spatio-temporal representation framework for visual navigation that improves feature encoding by modeling spatial and temporal structures, leading to better navigation performance.

Contribution

It proposes a unified framework with a spatio-temporal fusion module that enhances visual encoding for robotic navigation, incorporating graph reasoning and dynamic temporal modeling.

Findings

Improved navigation accuracy over baseline methods.

Enhanced visual feature representation through spatio-temporal fusion.

Generalizable backbone for goal-conditioned control.

Abstract

Visual navigation requires the robot to reach a specified goal such as an image, based on a sequence of first-person visual observations. While recent learning-based approaches have made significant progress, they often focus on improving policy heads or decision strategies while relying on simplistic feature encoders and temporal pooling to represent visual input. This leads to the loss of fine-grained spatial and temporal structure, ultimately limiting accurate action prediction and progress estimation. In this paper, we propose a unified spatio-temporal representation framework that enhances visual encoding for robotic navigation. Our approach extracts features from both image sequences and goal observations, and fuses them using the designed spatio-temporal fusion module. This module performs spatial graph reasoning within each frame and models temporal dynamics using a hybrid temporal shift module combined with multi-resolution difference-aware convolution. Experimental results demonstrate that our approach consistently improves navigation performance and offers a generalizable visual backbone for goal-conditioned control. Code is available at \href{https://github.com/hren20/STRNet}{https://github.com/hren20/STRNet}.