When Depth Hurts: Reliability-Aware Geometry Distillation for Depth-Free RGB-D Salient Object Detection

• URL: http://arxiv.org/abs/2609.03378v1
• Date: Thu, 03 Sep 2026 05:29:10 GMT
• Title: When Depth Hurts: Reliability-Aware Geometry Distillation for Depth-Free RGB-D Salient Object Detection
• Abstract summary: 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.<n>We propose method, a reliability-aware geometry distillation framework developed for RGB-D SOD benchmarks without using dataset-provided depth during training or inference.<n>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.
• Score: 39.23633029110609
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• 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.

Related papers

• GeomPrompt: Geometric Prompt Learning for RGB-D Semantic Segmentation Under Missing and Degraded Depth [0.0]
  GeomPrompt is a lightweight adaptation module that synthesizes a task-driven geometric prompt from RGB alone.<n>GeomPrompt-Recovery compensates for degraded depth by predicting the fourth channel correction relevant for the frozen segmenter.
  arXiv  Detail & Related papers  (2026-04-13T15:01:22Z)
• DepthMatch: Semi-Supervised RGB-D Scene Parsing through Depth-Guided Regularization [43.974708665104565]
  We introduce DepthMatch, a semi-supervised learning framework that is specifically designed for RGB-D scene parsing.<n>We propose complementary patch mix-up augmentation to explore the latent relationships between texture and spatial features in RGB-D image pairs.<n>We also design a lightweight spatial prior injector to replace traditional complex fusion modules, improving the efficiency of heterogeneous feature fusion.
  arXiv  Detail & Related papers  (2025-05-26T14:26:31Z)
• RGB-based Category-level Object Pose Estimation via Decoupled Metric Scale Recovery [72.13154206106259]
  We propose a novel pipeline that decouples the 6D pose and size estimation to mitigate the influence of imperfect scales on rigid transformations. Specifically, we leverage a pre-trained monocular estimator to extract local geometric information. A separate branch is designed to directly recover the metric scale of the object based on category-level statistics.
  arXiv  Detail & Related papers  (2023-09-19T02:20:26Z)
• Robust RGB-D Fusion for Saliency Detection [13.705088021517568]
  We propose a robust RGB-D fusion method that benefits from layer-wise and trident spatial, attention mechanisms. Our experiments on five benchmark datasets demonstrate that the proposed fusion method performs consistently better than the state-of-the-art fusion alternatives.
  arXiv  Detail & Related papers  (2022-08-02T21:23:00Z)
• Pyramidal Attention for Saliency Detection [30.554118525502115]
  This paper exploits only RGB images, estimates depth from RGB, and leverages the intermediate depth features. We employ a pyramidal attention structure to extract multi-level convolutional-transformer features to process initial stage representations. We report significantly improved performance against 21 and 40 state-of-the-art SOD methods on eight RGB and RGB-D datasets.
  arXiv  Detail & Related papers  (2022-04-14T06:57:46Z)
• Boosting RGB-D Saliency Detection by Leveraging Unlabeled RGB Images [89.81919625224103]
  Training deep models for RGB-D salient object detection (SOD) often requires a large number of labeled RGB-D images. We present a Dual-Semi RGB-D Salient Object Detection Network (DS-Net) to leverage unlabeled RGB images for boosting RGB-D saliency detection.
  arXiv  Detail & Related papers  (2022-01-01T03:02:27Z)
• Data-Level Recombination and Lightweight Fusion Scheme for RGB-D Salient Object Detection [73.31632581915201]
  We propose a novel data-level recombination strategy to fuse RGB with D (depth) before deep feature extraction. A newly lightweight designed triple-stream network is applied over these novel formulated data to achieve an optimal channel-wise complementary fusion status between the RGB and D.
  arXiv  Detail & Related papers  (2020-08-07T10:13:05Z)
• Depth Quality Aware Salient Object Detection [52.618404186447165]
  This paper attempts to integrate a novel depth quality aware into the classic bi-stream structure, aiming to assess the depth quality before conducting the selective RGB-D fusion. Compared with the SOTA bi-stream methods, the major highlight of our method is its ability to lessen the importance of those low-quality, no-contribution, or even negative-contribution D regions during the RGB-D fusion.
  arXiv  Detail & Related papers  (2020-08-07T09:54:39Z)
• Bi-directional Cross-Modality Feature Propagation with Separation-and-Aggregation Gate for RGB-D Semantic Segmentation [59.94819184452694]
  Depth information has proven to be a useful cue in the semantic segmentation of RGBD images for providing a geometric counterpart to the RGB representation. Most existing works simply assume that depth measurements are accurate and well-aligned with the RGB pixels and models the problem as a cross-modal feature fusion. In this paper, we propose a unified and efficient Crossmodality Guided to not only effectively recalibrate RGB feature responses, but also to distill accurate depth information via multiple stages and aggregate the two recalibrated representations alternatively.
  arXiv  Detail & Related papers  (2020-07-17T18:35:24Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.