Method is proposed, a reliability-aware geometry distillation framework developed for RGB-D SOD benchmarks without using dataset-provided depth during training or inference, and achieves the best or tied-best result in 26 of 36 metric-dataset comparisons against ten recent RGB-D SOD methods.
Abstract
Depth can resolve appearance ambiguity in RGB-D salient object detection (SOD), yet sensor depth is not uniformly reliable. Missing regions, blurred boundaries, and structural artifacts can propagate through multimodal fusion and make an RGB-D detector less accurate than its RGB-only counterpart. Existing quality-aware approaches regulate observed depth but remain dependent on the same potentially defective modality. We propose \method, a reliability-aware geometry distillation framework developed for RGB-D SOD benchmarks without using dataset-provided depth during training or inference. A frozen Depth Anything V2 model serves only as a training-time teacher, transferring dense relative geometry, hierarchical spatial attention, and boundary structure to a compact edge-aware geometry branch. Pooled bidirectional interaction aligns geometry with appearance, and a pixel-wise reliability estimator selectively injects geometry that is compatible with the current RGB representation. The teacher is removed after training, leaving an RGB-only inference network. Trained on 2,985 RGB-mask pairs, \method{} achieves the best or tied-best result in 26 of 36 metric-dataset comparisons against ten recent RGB-D SOD methods, including a 13.4\% relative MAE reduction on ReDWeb-S. When retrained on DUTS-TR, it also improves the strongest prior $F$-measure by 4.2\% on PASCAL-S, showing that the distilled geometry transfers beyond a particular sensor or dataset domain. Code will be released upon publication.
An end-to-end 2D-to-3D pipeline that consolidates 2D semantic features for 3D segmentation and delivers the strongest zero-shot transfer to SUN RGB-D and SceneNN is developed.
Xu-Ying Huang, Swithinraj Moses Daniel, Sicong Pan et al.· 0 citations
A Boundary-Guided Dual-Perspective Cross-Modal Fusion Network (BDPNet) is proposed to explicitly preserve shallow geometric structures and decouple deep semantic fusion into macroscopic and microscopic perspectives.
Hu Lin, Zhi-Wei Fu, Xiu-Mei Chen et al.· Remote Sensing· 0 citations
Recent event-based depth estimation methods successfully transfer geometric priors from vision foundation models via cross-modal distillation. However, their reliance on synchronized RGB-event pairs or depth annotations during training severely restricts practical deployment. To overcome this bottleneck, we propose SFE...
RGB-Thermal (RGB-T) salient object detection leverages complementary cues from visible and thermal modalities to improve robustness in challenging environments. However, in real-world scenarios, the reliability of each modality is inherently unstable: RGB images degrade under low illumination, motion blur, and noise, w...
Hong-Bo Gao, Zheng-Yu Li, Xue-Ru Nie et al.· 2 citations· ⚡1
DGQ-YOLO, an RGB-dominant detector that introduces pseudo-depth at the backbone and neck at the backbone and neck is proposed, which improves pseudo-depth-assisted aerial detection while retaining detector-level efficiency, although online throughput remains constrained by pseudo-depth generation.
Xin-Xing Hou, Wenlin Liu· Journal of King Saud Univers...· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.