Object Segmentation by Mining Cross-Modal Semantics

TL;DR

This paper introduces XMSNet, a novel multi-sensor object segmentation network that effectively mines cross-modal semantics to improve fusion, decoding, and robustness across various datasets and challenging tasks.

Contribution

The paper proposes a new network, XMSNet, which explicitly models shared and specific features across modalities to enhance segmentation accuracy and robustness.

Findings

Outperforms existing methods on eleven datasets.

Effective in salient and camouflage object segmentation.

Demonstrates robustness to sensor noise and calibration errors.

Abstract

Multi-sensor clues have shown promise for object segmentation, but inherent noise in each sensor, as well as the calibration error in practice, may bias the segmentation accuracy. In this paper, we propose a novel approach by mining the Cross-Modal Semantics to guide the fusion and decoding of multimodal features, with the aim of controlling the modal contribution based on relative entropy. We explore semantics among the multimodal inputs in two aspects: the modality-shared consistency and the modality-specific variation. Specifically, we propose a novel network, termed XMSNet, consisting of (1) all-round attentive fusion (AF), (2) coarse-to-fine decoder (CFD), and (3) cross-layer self-supervision. On the one hand, the AF block explicitly dissociates the shared and specific representation and learns to weight the modal contribution by adjusting the \textit{proportion, region,} and \textit{pattern}, depending upon the quality. On the other hand, our CFD initially decodes the shared feature and then refines the output through specificity-aware querying. Further, we enforce semantic consistency across the decoding layers to enable interaction across network hierarchies, improving feature discriminability. Exhaustive comparison on eleven datasets with depth or thermal clues, and on two challenging tasks, namely salient and camouflage object segmentation, validate our effectiveness in terms of both performance and robustness. The source code is publicly available at https://github.com/Zongwei97/XMSNet.