Pith. sign in

REVIEW 6 cited by

DFormer: Rethinking RGBD Representation Learning for Semantic Segmentation

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2309.09668 v2 pith:DU3Q5YQD submitted 2023-09-18 cs.CV

Bowen Yin , Xuying Zhang , Zhongyu Li , Li Liu , Ming-Ming Cheng , Qibin Hou This is my paper · ORCID
classification cs.CV
keywords rgb-ddformersegmentationpretrainedsemantictasksbackbonedatasets
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

We present DFormer, a novel RGB-D pretraining framework to learn transferable representations for RGB-D segmentation tasks. DFormer has two new key innovations: 1) Unlike previous works that encode RGB-D information with RGB pretrained backbone, we pretrain the backbone using image-depth pairs from ImageNet-1K, and hence the DFormer is endowed with the capacity to encode RGB-D representations; 2) DFormer comprises a sequence of RGB-D blocks, which are tailored for encoding both RGB and depth information through a novel building block design. DFormer avoids the mismatched encoding of the 3D geometry relationships in depth maps by RGB pretrained backbones, which widely lies in existing methods but has not been resolved. We finetune the pretrained DFormer on two popular RGB-D tasks, i.e., RGB-D semantic segmentation and RGB-D salient object detection, with a lightweight decoder head. Experimental results show that our DFormer achieves new state-of-the-art performance on these two tasks with less than half of the computational cost of the current best methods on two RGB-D semantic segmentation datasets and five RGB-D salient object detection datasets. Our code is available at: https://github.com/VCIP-RGBD/DFormer.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Semantics-aware Predictive Inspection Path Planning

    cs.RO 2025-06 conditional novelty 6.0 of 10

    A semantics-aware inspection planner that predicts repeated structures in unseen space reduces mission time by 12 to 19 percent in real ballast tanks while maintaining coverage.

  2. BiXFormer: A Robust Framework for Maximizing Modality Effectiveness in Multi-Modal Semantic Segmentation

    cs.CV 2025-06 conditional novelty 6.0 of 10

    A multi-modal semantic segmentation framework that processes RGB and non-RGB sensors separately, matches labels in two stages, and aligns cross-modal queries with a VAE refiner.

  3. Toward Reliable RGB-D Semantic Segmentation: Handling Missing Modalities via Condition Dropout

    cs.CV 2026-07 conditional novelty 5.0 of 10

    Condition Dropout freezes the original RGB-D encoder, trains a copied encoder with random modality dropout, and cuts the mIoU drop under missing inputs from about -25 to about -10 on NYUv2 and SUN RGB-D.

  4. HDBFormer: Efficient RGB-D Semantic Segmentation with A Heterogeneous Dual-Branch Framework

    cs.CV 2025-04 conditional novelty 5.0 of 10

    A heterogeneous dual-branch RGB-D segmentation network with a lightweight depth encoder and a transformer-plus-large-kernel fusion module matches or improves state-of-the-art mIoU on NYUDepthv2 and SUN-RGBD.

  5. Ultra-High Resolution Segmentation via Boundary-Enhanced Patch-Merging Transformer

    cs.CV 2024-12 conditional novelty 5.0 of 10

    A patch-merging transformer with a boundary-enhanced module improves state-of-the-art ultra-high resolution segmentation accuracy across five benchmarks with comparable memory use.

  6. Efficient Segment Anything with Depth-Aware Fusion and Limited Training Data

    cs.CV 2026-02 reject novelty 4.0 of 10

    Adding monocular depth to EfficientViT-SAM improves point-prompted segmentation at 3 and 5 clicks after fine-tuning on 11.2k images, but universal gains and data-efficiency are not established.

Pith tools