Pith. sign in

REVIEW 4 major objections 10 minor 43 references

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

T0 review · 4 major / 10 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Bidirectional Interaction Mamba (BIM) claims to preserve complete cross-task information in dense prediction at linear interaction complexity, and reports state-of-the-art numbers on NYUD-V2 and PASCAL-Context.

desk verdict A genuinely new Mamba decoder module with clear ablations, but the SOTA claim rests on single-run, test-set-selected numbers. read the letter →

arxiv 2508.20376 v1 pith:VQGWGDEH submitted 2025-08-28 cs.CV

Mang Cao , Sanping Zhou , Yizhe Li , Ye Deng , Wenli Huang , Le Wang This is my paper · ORCID
classification cs.CV
keywords multi-taskdensepredictionMambaBidirectionalInteractionScancross-taskstatespacemodelmulti-scalesemanticsegmentationdepthestimation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to settle the trade-off between interaction completeness and computational cost in multi-task dense prediction. It proposes the Bidirectional Interaction Mamba (BIM), a decoder built on Mamba's state-space model, whose Bidirectional Interaction Scan (BI-Scan) serializes task features in both task-first and position-first orders so that every task's full spatial information participates in cross-task exchange while interaction complexity stays linear in the number of tasks. A second mechanism, Multi-Scale Scan (MS-Scan), splits features across channels and scans them at several window scales to capture scene structure at multiple granularities. The paper reports state-of-the-art results on NYUD-V2 and PASCAL-Context, including semantic segmentation mIoU of 57.40 on NYUD-V2 versus 55.82 for the prior best method, at 547 GFLOPs. If these results hold, BIM shows that a dense-prediction decoder can exchange complete task information across all tasks at linear cost, removing the main obstacle to scaling cross-task interaction to many tasks.

What carries the argument

The two load-bearing mechanisms are BI-Scan and MS-Scan. BI-Scan is a bidirectional sequence construction for cross-task interaction: task-first mode serializes each task's tokens by a spatial scan pattern and concatenates the resulting subsequences along the task order; position-first mode regroups tokens that share a spatial position across tasks into length-$T$ subsequences and concatenates those along the scan pattern. Both directions are passed through the selective state-space model of Mamba, and the backward pass is reversed along the task dimension before concatenation. MS-Scan partitions the feature channels into branches, tokenizes each branch at a different window scale (for example $\{1,4\}$), runs SS2D four-way scanning in each branch, and fuses the branches back by inverse windowing and channel concatenation. The paper's argument is that this combination preserves every task's full spatial information during interaction (no information is compressed through a bridge representation) while keeping interaction complexity $O(n)$, and supplies multi-scale scene structure that different tasks can weight through gating.

What would settle it

Re-run BIM and MTMamba on NYUD-V2 and PASCAL-Context under the paper's exact training protocol with, say, five random seeds each, and compare the BIM-minus-MTMamba deltas (for example +1.58 semseg mIoU and $-0.033$ depth RMSE on NYUD-V2) against the seed-level standard deviation. If those deltas do not exceed the seed noise, the claimed superiority over the previous best method is not established.

Watch

Extended reading notes

Core claim

The central claim is that cross-task interaction in dense prediction does not require choosing between information integrity and linear complexity. BIM's BI-Scan takes each task's normalized feature map, splits it along channels, and builds two complementary sequences: task-first, where each task's full spatial scan is concatenated along the task order into one long sequence of length $T \times (H \times W)$, and position-first, where spatially aligned tokens from all tasks are gathered into $H \times W$ subsequences of length $T$. Both are processed by the selective state-space model from Mamba, giving each task access to the complete information of the other tasks while the cost grows only linearly with $T$. MS-Scan adds multi-granularity scene modeling by partitioning channels and scanning non-overlapping windows at multiple scales, with SS2D-style four-way scans per branch. On NYUD-V2 the paper reports Semseg mIoU 57.40, Depth RMSE 0.4733, Normal mErr 18.55, Boundary odsF 78.72, and on PASCAL-Context Semseg mIoU 81.25, Parsing mIoU 73.15, Saliency maxF 84.14, Normal mErr 14.13, Boundary odsF 79.05, with 547 GFLOPs; the ablation tables attribute the gains to the bidirectional scan and to the multi-scale scan.

Load-bearing premise

The load-bearing premise is that the reported accuracy differences between BIM and the previous best method are larger than run-to-run training variation, since every configuration in the comparison tables is evaluated once, with no seeds, standard deviations, or significance tests reported.

Editorial extensions

If this is right

  • If BIM is correct, the completeness-efficiency trade-off in cross-task interaction is not inherent: a decoder can give every task access to every other task's full features at linear interaction cost.
  • The two scan modes are complementary: combining task-first and position-first modes outperforms either alone, with negligible added computation, so future multi-task decoders can treat the two orderings as paired modules.
  • Multi-scale scanning improves Mamba-based image modeling: the $\{1,4\}$ scan-scale setting yields the largest multi-task gain, indicating that granularity diversity, not just longer sequences, matters for dense prediction.
  • The design scales predictably with task count: adding a fourth task on NYUD-V2 costs the same +86 GFLOPs and +42M parameters as adding the third task, consistent with the claimed linear complexity.
  • Task-order robustness means the method does not need per-dataset task-ordering search: several fixed orders perform similarly, and only random order degrades somewhat.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Going beyond the paper: because the two scan orderings are defined purely by serialization, the same BI-Scan structure could be paired with any linear sequence encoder, not just Mamba's selective state-space model; a testable extension is whether the gains persist with a different linear sequence layer, which would separate the scan design from the state-space backbone.
  • Going beyond the paper: the reported per-task cost that stays constant as tasks are added suggests BIM is a plausible decoder for many-task settings (well beyond four tasks) where pairwise or transformer interactions become prohibitive; the paper does not run that regime, but it is a direct quantitative prediction of the architecture.
  • Going beyond the paper: the ablation shows task-first and position-first scans are complementary, so a natural next experiment is applying BI-Scan at multiple decoder depths or interleaving it with MS-Scan more than once per MFR block to see whether the current single application is the saturation point.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 10 minor

Summary. The paper proposes BIM, a Mamba-based decoder for multi-task dense prediction. The decoder replaces MTMamba's cross-task scan with BI-Scan, which concatenates task features into long sequences in task-first and position-first modes and processes them bidirectionally with an SSM, and with MS-Scan, which tokenizes at multiple scales before SSM processing. Experiments on NYUD-V2 and PASCAL-Context report improvements over MTMamba and other baselines (e.g., NYUD-V2 semantic segmentation mIoU 57.40 vs. 55.82; PASCAL-Context parsing mIoU 73.15 vs. 72.62), and ablations attribute gains to both scan mechanisms. The paper's central claim is that BIM achieves state-of-the-art multi-task dense prediction while retaining linear interaction complexity.

Significance. The design is well specified and the component ablations are internally consistent, which makes the proposal credible as a method. The BI-Scan idea of injecting task-order and position-order serializations into a linear-complexity SSM is a useful and clearly explained contribution, and MS-Scan is a sensible adaptation of Mamba to multi-scale visual structure. If the numbers are reproducible, the NYUD-V2 improvements in semantic segmentation and depth are nontrivial. The availability of code is also a strength. However, the empirical evidence for the 'state-of-the-art' claim is currently thin: every configuration is evaluated once, several PASCAL-Context deltas are tiny or zero, and hyperparameters are selected on the test set. These issues are fixable but need to be addressed before the central claim can be accepted.

major comments (4)
  1. [Section 4.2, Tables 1-2] The central SOTA claim is based on a single evaluation per configuration with no seeds, standard deviations, or significance tests. On PASCAL-Context, Table 2 shows saliency exactly tying MTMamba at 84.14, normal estimation improving by only 0.01 mErr, and boundary/semseg/parsing changing by 0.25, 0.14, and 0.53 mIoU, respectively; these deltas are within the range of run-to-run variation typical for dense prediction benchmarks. The NYUD-V2 headline deltas are larger, but they are single runs compared against published MTMamba numbers rather than a controlled reimplementation. Please report mean and standard deviation over at least three seeds (or an appropriate significance test) and clarify whether the MTMamba baseline was retrained under exactly the same data pipeline.
  2. [Section 4.3, Tables 5-7] The final configuration is selected using NYUD-V2 test-set numbers. Table 5 chooses task order S-D-N-B, Table 6 chooses scan scale {1,4}, and Table 7 chooses Type 2, all based on the test metrics reported in those tables; the final Table 1 numbers are then produced by the selected configuration. Selection on the test set can inflate the reported gains and makes the headline improvements optimistic. Please move configuration selection to a held-out validation split, or report the test-set results of all configurations and disclose the selection protocol.
  3. [Section 3.4, Eq. (10)] The loss weights lambda_t in Eq. (10) are never specified in the implementation details of Section 4.1. Multi-task performance is sensitive to loss weighting, so the reader cannot reproduce the reported numbers without this information. Please state the values used for all tasks on both datasets.
  4. [Section 4.2, Table 1] The comparisons in Tables 1 and 2 mix numbers taken from prior publications with numbers produced by the authors' own pipeline. For the central comparison against MTMamba, the difference between published numbers and a controlled reimplementation is confounded with the architectural contribution. Please either retrain all relevant baselines with the same augmentation, iteration count, learning-rate schedule, and seeds, or clearly restrict the claim to 'our implementation configuration.'
minor comments (10)
  1. [Section 1, Contributions] There is a typo in the first contribution bullet: 'Bidricational' should be 'Bidirectional.'
  2. [Figure 1] The labels 'Intercation Complexity' should read 'Interaction Complexity.'
  3. [Figure 3] The label 'Flap & Concat' appears to be a typo; it should likely read 'Flatten & Concat.'
  4. [Section 3.3, Eq. (3)] The sentence introducing F_{t,N}^s contains a duplicated 'is': 'F_{t,N}^s is is derived' should read 'F_{t,N}^s is derived.'
  5. [Section 3.3, MS-Scan] The text says multi-scale scanning is introduced 'without increasing computational complexity,' but Table 3 reports BIM at 547 GFLOPs versus MTMamba's 541 GFLOPs. Please either provide a precise FLOP calculation or soften the statement to 'with small additional cost.'
  6. [Section 3.3, MS-Scan] The description of window tokenization assumes that the channel dimension C is divisible by the number of branches N; please state this assumption explicitly.
  7. [Table 3] DBIM is included in Table 3 but is only described in the supplementary material; either describe it in the main text or remove it from the main table.
  8. [Figure 6] The bar charts in Figure 6 do not show numerical values on the y-axis, making the claimed differences hard to verify; please add value labels or a table.
  9. [Section 4.1, Implementation Details] The phrase 'preliminary decoder has an output channel number of 768' is unclear; presumably 'initial decoder channel count' is meant.
  10. [Supplementary, Figure 8] The caption describes a 'quantitative visual analysis,' but the figure is qualitative; please rephrase.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: BIM's contribution is an architectural mechanism evaluated on external benchmarks, not a derivation that reduces to its inputs.

full rationale

The paper's central claim is that the proposed BIM decoder, with BI-Scan and MS-Scan, improves multi-task dense prediction on NYUD-V2 and PASCAL-Context. No step in the derivation chain defines the proposed mechanism in terms of the reported outcome. BI-Scan is defined by scanning and serialization operations (Eqs. 5-6), and MS-Scan by channel splitting, window tokenization, SS2D scanning, and inverse fusion (Eqs. 7-9); the reported metrics come from training these modules on external benchmarks. The comparison against MTMamba, VMamba, and other methods is empirical benchmarking, not self-citation used to justify a premise. Ablations such as Table 4 compare design variants and do not treat a fitted parameter as a prediction: no parameter is fitted to a subset of test data and then reported as a predicted result. The reliance on MTMamba as a baseline and the use of SS2D as a component are standard external references, and neither imports an unverified uniqueness theorem. Potential concerns about single-run evaluation and small deltas are validity or robustness issues, not circularity, and do not affect the circularity score.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The central empirical claim rests on three domain assumptions about Mamba state-space propagation, multi-scale tokenization alignment, and fair comparison protocol. The only fitted design choices are the MS-Scan scale set and BI-Scan task order, both chosen on the NYUD-V2 test set. The paper introduces no new physical or mathematical entities, so invented_entities is empty.

free parameters (2)
  • MS-Scan scale set = {1,4} for all MFR blocks
    Selected by test-set ablation on NYUD-V2 (Table 6); other scale sets give similar scores, so the final choice is only weakly constrained by validation.
  • BI-Scan task order = S-D-N-B (Semseg, Depth, Normal, Boundary)
    Chosen among fixed orders on NYUD-V2 test set (Table 5); random order is worse, while fixed orders differ only slightly.
assumptions (3)
  • domain assumption Mamba/SSM sequence modeling with four-way SS2D scans can serve as a linear-complexity surrogate for global cross-task attention.
    The design assumes selective state-space layers propagate task information across concatenated task sequences without losing the claimed 'information integrity'; no explicit bound or proof is given in Section 3.3, BI-Scan.
  • domain assumption Multi-scale window tokenization (MS-Scan) preserves enough spatial alignment after inverse windowing to fuse branches by simple channel concatenation.
    The reconstruction step S^-1(W^-1_i(y_i)) in Eqs. 7-9 assumes splitting and merging patches loses no task-critical alignment; this is not formally demonstrated.
  • domain assumption Baseline MTMamba's training recipe and evaluation protocol are fair and transferable to BIM.
    The comparison relies on matching training iterations, batch sizes, and augmentation; the paper follows 'common practice [16,35]' but does not state whether BIM's hyperparameters were tuned separately (Section 4.1).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Enhancing Mamba Decoder with Bidirectional Interaction in Multi-Task Dense Prediction." pith.science (2026). https://pith.science/paper/VQGWGDEH

@misc{pith2026250820376,
  author       = {Pith},
  title        = {Pith review of: Enhancing Mamba Decoder with Bidirectional Interaction in Multi-Task Dense Prediction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VQGWGDEH}},
  note         = {Machine review of arXiv:2508.20376}
}
read the original abstract

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.

Figures

Figures reproduced from arXiv: 2508.20376 by the authors.

Figure 1
Figure 1. Illustration of different cross-task interaction strategies. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Framework of the proposed BIM for multi-task dense prediction. (a) Overview of BIM, illustrating with depth estimation and [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Framework of BI-Scan. It consists of two modules, [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Instructions for MS-Scan. It consists of three distinct op [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Qualitative comparison with the best-performing method on Pascal-Context. Our method generates better details. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Ablation study on Bidirectional Scan (left) and two scan [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Comparison of MS-Scan and DMS-Scan. B. More Ablation Studies Effect of scan mode order in BI-Scan. We performed an ablation study to assess the impact of the scanning order in BI-Scan, containing two sequences: TF → PF (Task-First mode then Position-First mode) and PF …
Figure 8
Figure 8. Figure 8: Effect of BI-Scan and MS-Scan on attention patterns. [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]
Figure 11
Figure 11. Figure 11: More qualitative comparison on Pascal-Context. [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]
Figure 12
Figure 12. Figure 12: More qualitative comparison on Pascal-Context. [PITH_FULL_IMAGE:figures/full_fig_p012_12.png]
Figure 9
Figure 9. Figure 9: Effect of bidirectional scan on attention patterns. [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]
Figure 10
Figure 10. Figure 10: More qualitative comparison on NYUD-v2. RGB Semseg Parsing Saliency Normal Boundry M T m a m b a O u r s G T [PITH_FULL_IMAGE:figures/full_fig_p012_10.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 26 canonical work pages

  1. [1]

    Mult: An end-to-end multitask learning transformer

    Deblina Bhattacharjee, Tong Zhang, Sabine S ¨usstrunk, and Mathieu Salzmann. Mult: An end-to-end multitask learning transformer. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 12031– 12041, 2022. 1, 2

    work page 2022

  2. [2]

    Exploring rela- tional context for multi-task dense prediction

    David Br ¨uggemann, Menelaos Kanakis, Anton Obukhov, Stamatios Georgoulis, and Luc Van Gool. Exploring rela- tional context for multi-task dense prediction. In Proceed- ings of the IEEE/CVF international conference on computer vision, pages 15869–15878, 2021. 6

    work page 2021

  3. [3]

    Detect what you can: Detecting and representing objects using holistic mod- els and body parts

    Xianjie Chen, Roozbeh Mottaghi, Xiaobai Liu, Sanja Fi- dler, Raquel Urtasun, and Alan Yuille. Detect what you can: Detecting and representing objects using holistic mod- els and body parts. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 1971–1978,

    1971

  4. [4]

    Gradnorm: Gradient normalization for adaptive loss balancing in deep multitask networks

    Zhao Chen, Vijay Badrinarayanan, Chen-Yu Lee, and An- drew Rabinovich. Gradnorm: Gradient normalization for adaptive loss balancing in deep multitask networks. In In- ternational conference on machine learning, pages 794–803. PMLR, 2018. 2

    work page 2018

  5. [5]

    Multi-task learning with deep neural networks: A survey

    Michael Crawshaw. Multi-task learning with deep neural networks: A survey. CoRR, abs/2009.09796, 2020. 1

    arXiv 2009

  6. [6]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, pages 248–255. Ieee, 2009. 6

    2009

  7. [7]

    Hungry hungry hippos: Towards language modeling with state space mod- els

    Daniel Y Fu, Tri Dao, Khaled Kamal Saab, Armin W Thomas, Atri Rudra, and Christopher Re. Hungry hungry hippos: Towards language modeling with state space mod- els. In The Eleventh International Conference on Learning Representations, 2023. 2

    work page 2023

  8. [8]

    Nddr-cnn: Layerwise feature fusing in multi-task cnns by neural discriminative dimensionality reduction

    Yuan Gao, Jiayi Ma, Mingbo Zhao, Wei Liu, and Alan L Yuille. Nddr-cnn: Layerwise feature fusing in multi-task cnns by neural discriminative dimensionality reduction. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 3205–3214, 2019. 1

    work page 2019

Show all 43 references
  1. [9]

    Mamba: Linear-time sequence mod- eling with selective state spaces, 2024

    Albert Gu and Tri Dao. Mamba: Linear-time sequence mod- eling with selective state spaces, 2024. 2, 4, 5

    2024

  2. [10]

    Efficiently mod- eling long sequences with structured state spaces

    Albert Gu, Karan Goel, and Christopher R´e. Efficiently mod- eling long sequences with structured state spaces. CoRR, abs/2111.00396, 2021. 2

    2021 arXiv

  3. [11]

    Combining recurrent, convolutional, and continuous-time models with linear state space layers

    Albert Gu, Isys Johnson, Karan Goel, Khaled Saab, Tri Dao, Atri Rudra, and Christopher R ´e. Combining recurrent, convolutional, and continuous-time models with linear state space layers. Advances in neural information processing sys- tems, 34:572–585, 2021. 2

    2021

  4. [12]

    Localmamba: Visual state space model with windowed selective scan

    Tao Huang, Xiaohuan Pei, Shan You, Fei Wang, Chen Qian, and Chang Xu. Localmamba: Visual state space model with windowed selective scan. arXiv preprint arXiv:2403.09338,

    arXiv

  5. [13]

    Quantifying task pri- ority for multi-task optimization

    Wooseong Jeong and Kuk-Jin Yoon. Quantifying task pri- ority for multi-task optimization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 363–372, 2024. 2

    2024

  6. [14]

    Multi-task learning using uncertainty to weigh losses for scene geome- try and semantics

    Alex Kendall, Yarin Gal, and Roberto Cipolla. Multi-task learning using uncertainty to weigh losses for scene geome- try and semantics. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 7482–7491,

  7. [15]

    Multi-task learning with 3d-aware regulariza- tion

    Wei-Hong Li, Steven McDonagh, Ales Leonardis, and Hakan Bilen. Multi-task learning with 3d-aware regulariza- tion. In The Twelfth International Conference on Learning Representations, 2024. 2

    2024

  8. [16]

    MTMamba: Enhancing multi-task dense scene understanding by mamba-based de- coders

    Baijiong Lin, Weisen Jiang, Pengguang Chen, Yu Zhang, Shu Liu, and Ying-Cong Chen. MTMamba: Enhancing multi-task dense scene understanding by mamba-based de- coders. In European Conference on Computer Vision, 2024. 1, 2, 3, 6, 7

    2024

  9. [17]

    Vmamba: Visual state space model

    Yue Liu, Yunjie Tian, Yuzhong Zhao, Hongtian Yu, Lingxi Xie, Yaowei Wang, Qixiang Ye, and Yunfan Liu. Vmamba: Visual state space model. CoRR, abs/2401.10166, 2024. 3, 4, 5

    2024 arXiv

  10. [18]

    Swin transformer: Hierarchical vision transformer using shifted windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin transformer: Hierarchical vision transformer using shifted windows. In Proceedings of the IEEE/CVF international conference on computer vision, pages 10012–10022, 2021. 3, 6

    2021

  11. [19]

    Swiss army knife: Synergizing biases in knowledge from vision foundation models for multi-task learning

    Yuxiang Lu, Shengcao Cao, and Yu-Xiong Wang. Swiss army knife: Synergizing biases in knowledge from vision foundation models for multi-task learning. arXiv preprint arXiv:2410.14633, 2024. 7

    2024 arXiv

  12. [20]

    Attentive single-tasking of multiple tasks

    Kevis-Kokitsi Maninis, Ilija Radosavovic, and Iasonas Kokkinos. Attentive single-tasking of multiple tasks. InPro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 1851–1860, 2019. 6

    2019

  13. [21]

    Long range language modeling via gated state spaces

    Harsh Mehta, Ankit Gupta, Ashok Cutkosky, and Behnam Neyshabur. Long range language modeling via gated state spaces. arXiv preprint arXiv:2206.13947, 2022. 2

    2022 arXiv

  14. [22]

    Cross-stitch networks for multi-task learning

    Ishan Misra, Abhinav Shrivastava, Abhinav Gupta, and Mar- tial Hebert. Cross-stitch networks for multi-task learning. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3994–4003, 2016. 6

    2016

  15. [23]

    Multi- task learning as a bargaining game

    Aviv Navon, Aviv Shamsian, Idan Achituve, Haggai Maron, Kenji Kawaguchi, Gal Chechik, and Ethan Fetaya. Multi- task learning as a bargaining game. arXiv preprint arXiv:2202.01017, 2022. 2

    2022 arXiv

  16. [24]

    Efficient computation sharing for multi-task visual scene understanding

    Sara Shoouri, Mingyu Yang, Zichen Fan, and Hun-Seok Kim. Efficient computation sharing for multi-task visual scene understanding. In Proceedings of the IEEE/CVF In- ternational Conference on Computer Vision , pages 17130– 17141, 2023. 1, 2

    2023

  17. [25]

    Indoor segmentation and support inference from rgbd images

    Nathan Silberman, Derek Hoiem, Pushmeet Kohli, and Rob Fergus. Indoor segmentation and support inference from rgbd images. In Computer Vision–ECCV 2012: 12th Eu- ropean Conference on Computer Vision, Florence, Italy, Oc- tober 7-13, 2012, Proceedings, Part V 12 , pages 746–760...

    2012

  18. [26]

    Ema-net: Effi- cient multitask affinity learning for dense scene predictions

    Dimitrios Sinodinos and Narges Armanfard. Ema-net: Effi- cient multitask affinity learning for dense scene predictions. arXiv preprint arXiv:2401.11124, 2024. 1 9

    2024 arXiv

  19. [27]

    Jimmy T. H. Smith, Andrew Warrington, and Scott W. Lin- derman. Simplified state space layers for sequence modeling. CoRR, abs/2208.04933, 2022. 2

    2022 arXiv

  20. [28]

    Task switching network for multi-task learn- ing

    Guolei Sun, Thomas Probst, Danda Pani Paudel, Nikola Popovi´c, Menelaos Kanakis, Jagruti Patel, Dengxin Dai, and Luc Van Gool. Task switching network for multi-task learn- ing. In Proceedings of the IEEE/CVF international confer- ence on computer vision, pages 8291–8300, 2021. 1

    2021

  21. [29]

    Mti-net: Multi-scale task interaction networks for multi-task learning

    Simon Vandenhende, Stamatios Georgoulis, and Luc Van Gool. Mti-net: Multi-scale task interaction networks for multi-task learning. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part IV 16, pages 527–543. Springer, 2020. 2, 6

    2020

  22. [30]

    Multi-task learning for dense prediction tasks: A survey

    Simon Vandenhende, Stamatios Georgoulis, Wouter Van Gansbeke, Marc Proesmans, Dengxin Dai, and Luc Van Gool. Multi-task learning for dense prediction tasks: A survey. IEEE transactions on pattern analysis and machine intelligence, 44(7):3614–3633, 2021. 1

    2021

  23. [31]

    Pad-net: Multi-tasks guided prediction-and-distillation net- work for simultaneous depth estimation and scene parsing

    Dan Xu, Wanli Ouyang, Xiaogang Wang, and Nicu Sebe. Pad-net: Multi-tasks guided prediction-and-distillation net- work for simultaneous depth estimation and scene parsing. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 675–684, 2018. 1, 2, 6

    2018

  24. [32]

    Multi-task learning with multi-query trans- former for dense prediction

    Yangyang Xu, Xiangtai Li, Haobo Yuan, Yibo Yang, and Lefei Zhang. Multi-task learning with multi-query trans- former for dense prediction. IEEE Transactions on Circuits and Systems for Video Technology, 2023. 2, 6

    2023

  25. [33]

    Chenhongyi Yang, Zehui Chen, Miguel Espinosa, Linus Er- icsson, Zhenyu Wang, Jiaming Liu, and Elliot J. Crowley. Plainmamba: Improving non-hierarchical mamba in visual recognition. CoRR, abs/2403.17695, 2024. 3

    2024 arXiv

  26. [34]

    Adaptive stochastic gradient algorithm for black- box multi-objective learning

    Feiyang Ye, Yueming Lyu, Xuehao Wang, Yu Zhang, and Ivor Tsang. Adaptive stochastic gradient algorithm for black- box multi-objective learning. In The Twelfth International Conference on Learning Representations, 2024. 2

    2024

  27. [35]

    Inverted pyramid multi-task trans- former for dense scene understanding

    Hanrong Ye and Dan Xu. Inverted pyramid multi-task trans- former for dense scene understanding. In European Confer- ence on Computer Vision, pages 514–530. Springer, 2022. 1, 2, 6, 7

    2022

  28. [36]

    Taskprompter: Spatial-channel multi-task prompting for dense scene understanding

    Hanrong Ye and Dan Xu. Taskprompter: Spatial-channel multi-task prompting for dense scene understanding. In The Eleventh International Conference on Learning Representa- tions, 2023. 2, 6, 7

    2023

  29. [37]

    Invpt++: Inverted pyramid multi- task transformer for visual scene understanding.IEEE Trans- actions on Pattern Analysis and Machine Intelligence, 2024

    Hanrong Ye and Dan Xu. Invpt++: Inverted pyramid multi- task transformer for visual scene understanding.IEEE Trans- actions on Pattern Analysis and Machine Intelligence, 2024. 6, 7

    2024

  30. [38]

    Gradient surgery for multi-task learning

    Tianhe Yu, Saurabh Kumar, Abhishek Gupta, Sergey Levine, Karol Hausman, and Chelsea Finn. Gradient surgery for multi-task learning. Advances in Neural Information Pro- cessing Systems, 33:5824–5836, 2020. 2

    2020

  31. [39]

    Re- thinking of feature interaction for multi-task learning on dense prediction

    Jingdong Zhang, Jiayuan Fan, Peng Ye, Bo Zhang, Hancheng Ye, Baopu Li, Yancheng Cai, and Tao Chen. Re- thinking of feature interaction for multi-task learning on dense prediction. arXiv preprint arXiv:2312.13514 , 2023. 1, 2

    2023 arXiv

  32. [40]

    Pattern-affinitive propagation across depth, surface normal and semantic segmentation

    Zhenyu Zhang, Zhen Cui, Chunyan Xu, Yan Yan, Nicu Sebe, and Jian Yang. Pattern-affinitive propagation across depth, surface normal and semantic segmentation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4106–4115, 2019. 6

    2019

  33. [41]

    Rs-mamba for large remote sensing image dense prediction

    Sijie Zhao, Hao Chen, Xueliang Zhang, Pengfeng Xiao, Lei Bai, and Wanli Ouyang. Rs-mamba for large remote sensing image dense prediction. arXiv preprint arXiv:2404.02668 ,

    arXiv

  34. [42]

    Pattern-structure diffu- sion for multi-task learning

    Ling Zhou, Zhen Cui, Chunyan Xu, Zhenyu Zhang, Chaoqun Wang, Tong Zhang, and Jian Yang. Pattern-structure diffu- sion for multi-task learning. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 4514–4523, 2020. 6

    2020

  35. [43]

    Vision mamba: Efficient visual representation learning with bidirectional state space model

    Lianghui Zhu, Bencheng Liao, Qian Zhang, Xinlong Wang, Wenyu Liu, and Xinggang Wang. Vision mamba: Efficient visual representation learning with bidirectional state space model. In ICML, 2024. 2, 3 10 Enhancing Mamba Decoder with Bidirectional Interaction in Multi-Task Dense P...

    2024

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.