|
|
|
|
| LEADER |
01000caa a22002652c 4500 |
| 001 |
NLM368441865 |
| 003 |
DE-627 |
| 005 |
20250305193607.0 |
| 007 |
cr uuu---uuuuu |
| 008 |
240215s2024 xx |||||o 00| ||eng c |
| 024 |
7 |
|
|a 10.1109/TIP.2024.3364022
|2 doi
|
| 028 |
5 |
2 |
|a pubmed25n1227.xml
|
| 035 |
|
|
|a (DE-627)NLM368441865
|
| 035 |
|
|
|a (NLM)38354080
|
| 040 |
|
|
|a DE-627
|b ger
|c DE-627
|e rakwb
|
| 041 |
|
|
|a eng
|
| 100 |
1 |
|
|a Chen, Hao
|e verfasserin
|4 aut
|
| 245 |
1 |
0 |
|a Disentangled Cross-Modal Transformer for RGB-D Salient Object Detection and Beyond
|
| 264 |
|
1 |
|c 2024
|
| 336 |
|
|
|a Text
|b txt
|2 rdacontent
|
| 337 |
|
|
|a ƒaComputermedien
|b c
|2 rdamedia
|
| 338 |
|
|
|a ƒa Online-Ressource
|b cr
|2 rdacarrier
|
| 500 |
|
|
|a Date Revised 06.03.2024
|
| 500 |
|
|
|a published: Print-Electronic
|
| 500 |
|
|
|a Citation Status PubMed-not-MEDLINE
|
| 520 |
|
|
|a Previous multi-modal transformers for RGB-D salient object detection (SOD) generally directly connect all patches from two modalities to model cross-modal correlation and perform multi-modal combination without differentiation, which can lead to confusing and inefficient fusion. Instead, we disentangle the cross-modal complementarity from two views to reduce cross-modal fusion ambiguity: 1) Context disentanglement. We argue that modeling long-range dependencies across modalities as done before is uninformative due to the severe modality gap. Differently, we propose to disentangle the cross-modal complementary contexts to intra-modal self-attention to explore global complementary understanding, and spatial-aligned inter-modal attention to capture local cross-modal correlations, respectively. 2) Representation disentanglement. Unlike previous undifferentiated combination of cross-modal representations, we find that cross-modal cues complement each other by enhancing common discriminative regions and mutually supplement modal-specific highlights. On top of this, we divide the tokens into consistent and private ones in the channel dimension to disentangle the multi-modal integration path and explicitly boost two complementary ways. By progressively propagate this strategy across layers, the proposed Disentangled Feature Pyramid module (DFP) enables informative cross-modal cross-level integration and better fusion adaptivity. Comprehensive experiments on a large variety of public datasets verify the efficacy of our context and representation disentanglement and the consistent improvement over state-of-the-art models. Additionally, our cross-modal attention hierarchy can be plug-and-play for different backbone architectures (both transformer and CNN) and downstream tasks, and experiments on a CNN-based model and RGB-D semantic segmentation verify this generalization ability
|
| 650 |
|
4 |
|a Journal Article
|
| 700 |
1 |
|
|a Shen, Feihong
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Ding, Ding
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Deng, Yongjian
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Li, Chao
|e verfasserin
|4 aut
|
| 773 |
0 |
8 |
|i Enthalten in
|t IEEE transactions on image processing : a publication of the IEEE Signal Processing Society
|d 1992
|g 33(2024) vom: 14., Seite 1699-1709
|w (DE-627)NLM09821456X
|x 1941-0042
|7 nnas
|
| 773 |
1 |
8 |
|g volume:33
|g year:2024
|g day:14
|g pages:1699-1709
|
| 856 |
4 |
0 |
|u http://dx.doi.org/10.1109/TIP.2024.3364022
|3 Volltext
|
| 912 |
|
|
|a GBV_USEFLAG_A
|
| 912 |
|
|
|a SYSFLAG_A
|
| 912 |
|
|
|a GBV_NLM
|
| 912 |
|
|
|a GBV_ILN_350
|
| 951 |
|
|
|a AR
|
| 952 |
|
|
|d 33
|j 2024
|b 14
|h 1699-1709
|