Enhancing Mamba Decoder with Bidirectional Interaction in Multi-Task Dense Prediction

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cao, Mang, Zhou, Sanping, Li, Yizhe, Deng, Ye, Huang, Wenli, Wang, Le
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908507641479168
author Cao, Mang
Zhou, Sanping
Li, Yizhe
Deng, Ye
Huang, Wenli
Wang, Le
author_facet Cao, Mang
Zhou, Sanping
Li, Yizhe
Deng, Ye
Huang, Wenli
Wang, Le
contents Sufficient cross-task interaction is crucial for success in multi-task dense prediction. However, sufficient interaction often results in high computational complexity, forcing existing methods to face the trade-off between interaction completeness and computational efficiency. To address this limitation, this work proposes a Bidirectional Interaction Mamba (BIM), which incorporates novel scanning mechanisms to adapt the Mamba modeling approach for multi-task dense prediction. On the one hand, we introduce a novel Bidirectional Interaction Scan (BI-Scan) mechanism, which constructs task-specific representations as bidirectional sequences during interaction. By integrating task-first and position-first scanning modes within a unified linear complexity architecture, BI-Scan efficiently preserves critical cross-task information. On the other hand, we employ a Multi-Scale Scan~(MS-Scan) mechanism to achieve multi-granularity scene modeling. This design not only meets the diverse granularity requirements of various tasks but also enhances nuanced cross-task feature interactions. Extensive experiments on two challenging benchmarks, \emph{i.e.}, NYUD-V2 and PASCAL-Context, show the superiority of our BIM vs its state-of-the-art competitors.
format Preprint
id arxiv_https___arxiv_org_abs_2508_20376
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Enhancing Mamba Decoder with Bidirectional Interaction in Multi-Task Dense Prediction
Cao, Mang
Zhou, Sanping
Li, Yizhe
Deng, Ye
Huang, Wenli
Wang, Le
Computer Vision and Pattern Recognition
Sufficient cross-task interaction is crucial for success in multi-task dense prediction. However, sufficient interaction often results in high computational complexity, forcing existing methods to face the trade-off between interaction completeness and computational efficiency. To address this limitation, this work proposes a Bidirectional Interaction Mamba (BIM), which incorporates novel scanning mechanisms to adapt the Mamba modeling approach for multi-task dense prediction. On the one hand, we introduce a novel Bidirectional Interaction Scan (BI-Scan) mechanism, which constructs task-specific representations as bidirectional sequences during interaction. By integrating task-first and position-first scanning modes within a unified linear complexity architecture, BI-Scan efficiently preserves critical cross-task information. On the other hand, we employ a Multi-Scale Scan~(MS-Scan) mechanism to achieve multi-granularity scene modeling. This design not only meets the diverse granularity requirements of various tasks but also enhances nuanced cross-task feature interactions. Extensive experiments on two challenging benchmarks, \emph{i.e.}, NYUD-V2 and PASCAL-Context, show the superiority of our BIM vs its state-of-the-art competitors.
title Enhancing Mamba Decoder with Bidirectional Interaction in Multi-Task Dense Prediction
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.20376