CFPNet: Improving Lightweight ToF Depth Completion via Cross-zone Feature Propagation

TL;DR

CFPNet introduces a novel approach for lightweight ToF depth completion by effectively propagating features across zones using attention and large-kernel convolutions, significantly improving outside-zone depth accuracy.

Contribution

The paper proposes CFPNet with two novel modules for cross-zone feature propagation, addressing the limited FOV challenge in lightweight ToF sensors.

Findings

Achieves state-of-the-art depth completion performance on ZJU-L5 dataset.

Effective cross-zone feature propagation improves outside-zone depth accuracy.

Combines attention-based and large-kernel convolution modules for superior results.

Abstract

Depth completion using lightweight time-of-flight (ToF) depth sensors is attractive due to their low cost. However, lightweight ToF sensors usually have a limited field of view (FOV) compared with cameras. Thus, only pixels in the zone area of the image can be associated with depth signals. Previous methods fail to propagate depth features from the zone area to the outside-zone area effectively, thus suffering from degraded depth completion performance outside the zone. To this end, this paper proposes the CFPNet to achieve cross-zone feature propagation from the zone area to the outside-zone area with two novel modules. The first is a direct-attention-based propagation module (DAPM), which enforces direct cross-zone feature acquisition. The second is a large-kernel-based propagation module (LKPM), which realizes cross-zone feature propagation by utilizing convolution layers with kernel sizes up to 31. CFPNet achieves state-of-the-art (SOTA) depth completion performance by combining these two modules properly, as verified by extensive experimental results on the ZJU-L5 dataset. The code is available at https://github.com/denyingmxd/CFPNet.