Pith. sign in

REVIEW 3 major objections 5 minor 42 references

Box for Mask and Mask for Box: weak losses for multi-task partially supervised learning

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Weak label transfer between detection and segmentation improves COCO multi-task learning

desk verdict Useful MTPSL paper with a genuinely new Mask-for-Box refinement, but the headline COCO comparison is confounded by an un-ablated attention module added to the segmentation head. read the letter →

arxiv 2411.17536 v1 pith:G7ITJ6PW submitted 2024-11-26 cs.CV

Hoàng-Ân Lê , Paul Berg , Minh-Tan Pham This is my paper · ORCID
classification cs.CV
keywords multi-tasklearningpartiallysupervisedobjectdetectionsemanticsegmentationweaksupervisionpseudo-labelsBox-for-MaskMask-for-Box
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

The paper studies multi-task partially supervised learning, where each training image is annotated for exactly one task, either object detection or semantic segmentation, and asks whether the annotations of one task can be used to train the other task's head. It proposes two modules: Mask-for-Box, which refines the circumscribed rectangles of a segmentation mask's connected components using the model's own predicted boxes and trains detection with the localization loss only, and Box-for-Mask, which builds pseudo-masks from ground-truth boxes, including a box-filled mask for attention, a coarse mask for cross-entropy, and a triplet loss on embeddings, to train segmentation. Combined as BoMBo, the method improves both tasks over the multi-task baseline on COCO across all tested backbones, with detection mAP rising from 17.2 to 19.1 and segmentation IoU from 54.5 to 58.5 on ResNet50. On VOC the gain is mixed, improving both tasks only for the largest backbone. The paper's point is that noisy but weakly relevant targets from the other task's ground truth are usable if the noise is filtered by the network's own predictions and by attention modulation.

What carries the argument

The load-bearing mechanism is a pair of label-refinement modules that convert one task's ground truth into targets for the other task's head. Mask-for-Box uses a matching and refining algorithm: reference boxes are extracted as circumscribed rectangles of connected components in the ground-truth mask; predicted boxes with confidence above thresholds are matched to references, used to split multi-instance references, merge fragmented ones into a smallest covering box, or kept as is, and the surviving refined boxes are filtered by non-maximum suppression; only the localization loss is applied. Box-for-Mask generates two pseudo-masks from each ground-truth box, a box-shaped mask of filled categories that trains an auxiliary attention map via mean-squared error, and a coarse mask restricted to the box interior that trains the main cross-entropy loss, plus a triplet loss that anchors each box's mean embedding of correctly predicted pixels to a same-class key and pushes it away from a different-class key. The attention-modulated prediction $M = M_l \otimes \alpha$ is what lets the noisy pseudo-masks improve segmentation instead of degrading it.

What would settle it

Run BoMBo on two datasets with disjoint or only partially overlapping class sets, for example COCO detection with Cityscapes segmentation, keeping everything else fixed; if the gains on COCO disappear or turn negative, the shared-class-space assumption is what carries the result. Alternatively, replace the refined pseudo-boxes with the unrefined circumscribed rectangles of the ground-truth masks and check whether the detection margin on COCO collapses, isolating the contribution of the refining step.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that cross-task weak supervision works in multi-task partially supervised learning if the transfer is designed around what each annotation type can and cannot say. A ground-truth mask gives correct categories and an object's extent but not instance boundaries, so Mask-for-Box takes the circumscribed rectangles of connected components, which may be fragmented or contain many instances, and re-localizes them using the network's predicted boxes, splitting, merging, or keeping boxes according to overlap and a two-sided touch criterion, then supervises the detection head with only the localization loss, because adding classification loss hurts. A ground-truth box gives the object's category and an upper bound on its extent but no pixel-level boundary, so Box-for-Mask generates a box-shaped mask used to train an attention map, a coarse pseudo-mask used for cross-entropy, and a triplet loss pulling embeddings of pixels inside a box toward same-class embeddings and away from other classes; the segmentation prediction is the logits modulated by the learned attention. Combining the two losses with lambda equal to 2, BoMBo outperforms the multi-task baseline on COCO for detection and segmentation on every backbone tested, while on VOC it improves both tasks only for Swin-L and is mixed otherwise.

Load-bearing premise

The central assumption is that the detection and segmentation datasets share the same semantic class space and come from the same image domain, because the pseudo-targets inherit their category labels from the other task's ground truth; if the class sets differ, the weak losses would train wrong categories.

Editorial extensions

If this is right

  • In partially annotated settings where detection and segmentation datasets share a class space, one task's annotations can genuinely expand the other's training data, improving both tasks on COCO at one-eighth data scale.
  • Classification supervision on pseudo-boxes is counterproductive; the useful signal from mask-derived boxes is their localization, which should be trained with the localization loss only.
  • Box-derived pseudo-masks do not help when they directly drive the segmentation cross-entropy; they work only when mediated by an attention map trained on the box-shaped filling.
  • The gains transfer across backbones on COCO, from ResNet18 and ResNet50 to Swin-T, Swin-B, and Swin-L, suggesting the mechanism is not tied to one architecture.
  • Combining both modules requires balancing the tasks, with lambda set to 2, and the mixed VOC results indicate the balance interacts with dataset difficulty or domain closeness.

Reading between the lines

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

  • If the shared-class-space assumption holds, the same recipe could be tried for other task pairs with complementary annotation granularity, such as instance segmentation and depth, where one task's dense ground truth could be converted into box-like or mask-like targets for the other.
  • The localization-only finding hints that pseudo-targets should carry only the information the target task lacks, not the full supervision; a testable extension would be to decay the classification weight on pseudo-boxes gradually as the network's own classifiers mature.
  • The mixed VOC results suggest a concrete diagnostic: measure the per-class overlap between the detection and segmentation label distributions, and use that overlap to gate whether BoMBo modules are switched on for a given class.
  • Because Mask-for-Box improves most on miss and false-negative errors, it could be attached to a detector that is already strong at classification but weak at recall, using segmentation data to grow recall without touching classifier training.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper studies multi-task partially supervised learning (MTPSL) for object detection and semantic segmentation, where each training image is annotated for only one of the two tasks. It proposes two modules: Mask-for-Box (M4B), which refines circumscribed rectangles of ground-truth mask connected components using predicted boxes to generate localization-only pseudo-labels for the detector, and Box-for-Mask (B4M), which builds pseudo-masks from ground-truth boxes (box-filling plus GrabCut) and trains the segmentation head with a cross-entropy loss on the coarse pseudo-mask, an attention loss on box-shaped masks, and a triplet loss on embeddings. The two modules are combined into BoMBo and evaluated on VOC and COCO with ResNet and Swin backbones. The reported COCO results show consistent gains over the multi-task baseline, while VOC results are mixed.

Significance. If the reported gains are robust, the work is a useful contribution to exploiting heterogeneous partially annotated datasets in multi-task learning, building on prior MTPSL work. The paper contains several strengths: design-level ablations for each module (Tables 1, 3, 4), a TIDE error analysis for the detection component, experiments across multiple backbones, and a commitment to release source code and data splits. However, the main COCO comparison is confounded by an architectural change (an added attention module) that is not present in the baseline, and the VOC results are not consistently favorable, so the evidence for the central claim is currently partial.

major comments (3)
  1. [Section 6, Table 5] The comparison between BoMBo and the MTL baseline in Table 5 is confounded by an architectural change. Section 6 states that the network is 'kept unchanged from [16] except an additional attention module in the segmentation head.' The MTL baseline (from [16]) therefore lacks this attention module. The reported COCO segmentation improvements (e.g., ResNet50 from 54.535 to 58.466) could stem from the additional module alone, independently of the proposed cross-task losses. A necessary control experiment is an MTL baseline with the attention module added but without any Box-for-Mask or Mask-for-Box objectives. Without this control, the attribution of the COCO gains to the weak losses is not established. The same confound affects Table 4, where the '+Lα' row adds both the attention module and the attention loss relative to the baseline, so the effect of Lα is not isolated.
  2. [Table 5] The VOC results do not support the abstract's implication of generally favorable performance. On VOC, BoMBo improves both tasks only for SwinL; on ResNet50 both detection mAP and segmentation IoU decrease (55.174 to 54.885 and 75.658 to 74.861), and on SwinT/SwinB the segmentation IoU decreases despite detection gains. The paper notes an 'imbalance problem' in passing, but does not analyze why the combination helps on COCO and hurts on VOC, or why the effect is backbone-dependent. Since the central claim is that cross-task weak losses improve MTPSL, the VOC results are a load-bearing inconsistency that needs either a technical explanation (e.g., dataset-domain shift, class-space mismatch) or a tempered statement of the claim.
  3. [Section 4.1] All experimental numbers are reported from a single run with no error bars or multiple seeds. This is particularly concerning for the small differences in Table 5, such as VOC SwinB detection (58.267 vs 58.259) and the mixed VOC segmentation changes. Without variance estimates, the reader cannot assess whether any of the gains or losses are statistically meaningful. At least the main BoMBo versus MTL comparisons should be repeated with multiple seeds and the mean and standard deviation reported.
minor comments (5)
  1. [Abstract] There is a typo in the abstract: 'Bo x-for-Mask' should read 'Box-for-Mask'.
  2. [Eq. (4)] The summation notation 'k∈0...nB' is ambiguous; it should be 'k = 0, ..., nB−1'.
  3. [Section 3.3] The paper does not specify how the triplet-loss keys are stored and updated across batches, beyond stating that the EMA network is used. Clarify whether the keys are embeddings from the EMA network on previous-batch images and how the key bank is maintained.
  4. [Algorithm 1] The thresholds in the Mask-for-Box algorithm (IoU 0.6/0.8, confidence 0.1/0.4/0.5, NMS 0.4) are introduced without justification or sensitivity analysis. A small study varying these parameters would improve confidence in the method's robustness.
  5. [Section 3.2] The informal phrase 'a predicted box should have at least 2 sides touching those of a reference box' is only made precise in Algorithm 1's touch function. The main text should refer explicitly to the algorithm's definition to help the reader.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the weak losses are heuristic and evaluated on standard benchmarks; the only self-citation ([16]) is used as a baseline/implementation and is not load-bearing for the new claim.

full rationale

The paper's central claim is that cross-task weak losses (Mask-for-Box and Box-for-Mask) improve multi-task partially supervised learning. These losses are defined directly from ground-truth masks/boxes and network predictions (Eqs. 1-4); they are not defined in terms of the final detection mAP or segmentation IoU, and the final numbers are measured on standard VOC/COCO splits, so no fitted-to-test circularity is present. The stated limitation in Section 5 ('One limitation of the study is the assumptions of the same data domain and shared class space between the two tasks') is a generality caveat, not a circular step. The main self-citation is to [16], which supplies the MTPSL baseline, data split, and base architecture; while this is a same-author citation, the new weak-loss modules are independently ablated and compared, so the citation is not load-bearing for the central claim. One experimental concern, visible in Section 6 ('kept unchanged from [16] except an additional attention module in the segmentation head'), is that the BoMBo comparison in Table 5 adds an attention module absent from the MTL baseline; this could confound attribution of the COCO gains to the weak losses. That is an internal-validity/control issue, not a definitional circularity, because the weak-loss effect is not imposed by construction. The M4B refinement's use of the model's own predicted boxes to refine pseudo-boxes is a standard self-training/pseudo-labeling mechanism anchored in GT masks; it is self-referential in an engineering sense but not a logical reduction of the claimed result to its inputs. Overall, no equation or definition in the paper makes a prediction equivalent to its input by construction.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The method relies on several domain assumptions about label structure rather than on fitted physical constants. Hyperparameters such as the loss balance lambda and matching thresholds are chosen by hand without sensitivity analysis. The central evaluation is empirical and does not derive theoretical predictions.

free parameters (3)
  • loss balancing weight lambda = 2
    Chosen in Eq. 1 to balance detection and segmentation losses; no sensitivity analysis reported.
  • triplet margin gamma = 0.1
    Set in Eq. 4; no ablation on this value.
  • M4B thresholds (IoU 0.6/0.8, confidence 0.1/0.4/0.5, NMS 0.4) = Various (see Algorithm 1)
    Hand-chosen thresholds controlling matching and refinement in Algorithm 1; no systematic tuning discussed.
assumptions (4)
  • domain assumption Detection and segmentation datasets share the same class space and image domain.
    Stated in Section 5 as a limitation; if violated, pseudo-targets assign wrong categories.
  • domain assumption Ground truth boxes provide the upper limit of the object's extent, so pixels outside boxes are background.
    Used in Box-for-Mask to filter coarse masks (Section 3.3).
  • domain assumption Predicted boxes from a partially trained detector carry useful instance cues for splitting and merging.
    Core to Mask-for-Box refinement; if the detector is too noisy, refinement could add noise.
  • domain assumption The circumscribed rectangles of connected components in semantic masks provide reliable class and rough location priors.
    Basis of Mask-for-Box; the paper shows naively using them helps over baseline in most cases.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Box for Mask and Mask for Box: weak losses for multi-task partially supervised learning." pith.science (2026). https://pith.science/paper/G7ITJ6PW

@misc{pith2026241117536,
  author       = {Pith},
  title        = {Pith review of: Box for Mask and Mask for Box: weak losses for multi-task partially supervised learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/G7ITJ6PW}},
  note         = {Machine review of arXiv:2411.17536}
}
read the original abstract

Object detection and semantic segmentation are both scene understanding tasks yet they differ in data structure and information level. Object detection requires box coordinates for object instances while semantic segmentation requires pixel-wise class labels. Making use of one task's information to train the other would be beneficial for multi-task partially supervised learning where each training example is annotated only for a single task, having the potential to expand training sets with different-task datasets. This paper studies various weak losses for partially annotated data in combination with existing supervised losses. We propose Box-for-Mask and Mask-for-Box strategies, and their combination BoMBo, to distil necessary information from one task annotations to train the other. Ablation studies and experimental results on VOC and COCO datasets show favorable results for the proposed idea. Source code and data splits can be found at https://github.com/lhoangan/multas.

Figures

Figures reproduced from arXiv: 2411.17536 by the authors.

Figure 1
Figure 1. Multi-task partially supervised learning with two tasks, object detection (blue) and [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The Mask-for-Box module uses predicted boxes to refine the circumscribed rect [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The Box-for-Mask module generates pseudo-masks by filling the ground truth [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Qualitative results of refined boxes with magenta indicate the adding, yellow [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: The network architecture being used in the paper, redrawn from [ [PITH_FULL_IMAGE:figures/full_fig_p015_5.png]
Figure 6
Figure 6. Figure 6: Qualitative results, from left to right: predicted boxes, refined boxes by M4B, [PITH_FULL_IMAGE:figures/full_fig_p018_6.png]
Figure 7
Figure 7. Figure 7: Qualitative results, from left to right: predicted boxes, refined boxes by M4B, [PITH_FULL_IMAGE:figures/full_fig_p019_7.png]
Figure 8
Figure 8. Figure 8: Qualitative results of failure cases, from left to right: predicted boxes, refined [PITH_FULL_IMAGE:figures/full_fig_p020_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 42 canonical work pages

  1. [16]

    Data exploitation: multi-task learning of object detection and semantic segmentation on partially annotated data

    Hoàng-Ân Lê and Minh-Tan Pham. Data exploitation: multi-task learning of object detection and semantic segmentation on partially annotated data. In British Machine Vision Conference (BMVC), 2023

    work page 2023

  2. [1]

    Multiscale Combinatorial Grouping

    Pablo Arbeláez, Jordi Pont-Tuset, Jon Barron, Ferran Marques, and Jitendra Malik. Multiscale Combinatorial Grouping. In 2014 IEEE Conference on Computer Vision and Pattern Recognition, pages 328–335, 2014

    work page 2014

  3. [2]

    TIDE: A General Toolbox for Identifying Object Detection Errors

    Daniel Bolya, Sean Foley, James Hays, and Judy Hoffman. TIDE: A General Toolbox for Identifying Object Detection Errors. In ECCV, 2020

    work page 2020

  4. [3]

    Stochastic Filter Groups for Multi-Task CNNs: Learning Specialist and Gen- eralist Convolution Kernels

    Felix J S Bragman, Ryutaro Tanno, Sebastien Ourselin, Daniel C Alexander, and Jorge Cardoso. Stochastic Filter Groups for Multi-Task CNNs: Learning Specialist and Gen- eralist Convolution Kernels. InProceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), oct 2019

    work page 2019

  5. [4]

    Automated Search for Resource-Efficient Branched Multi-Task Networks

    David Brüggemann, Menelaos Kanakis, Stamatios Georgoulis, and Luc Van Gool. Automated Search for Resource-Efficient Branched Multi-Task Networks. In BMVC, 2020

    work page 2020

  6. [5]

    Just Pick a Sign: Optimizing Deep Multitask Mod- els with Gradient Sign Dropout

    Zhao Chen, Jiquan Ngiam, Yanping Huang, Thang Luong, Henrik Kretzschmar, Yun- ing Chai, and Dragomir Anguelov. Just Pick a Sign: Optimizing Deep Multitask Mod- els with Gradient Sign Dropout. In H Larochelle, M Ranzato, R Hadsell, M F Balcan, and H Lin, editors, Advances in Neural Information Processing Systems , volume 33, pages 2039–2050. Curran Associa...

    work page 2020

  7. [6]

    A Multi-Task Mean Teacher for Semi-Supervised Shadow Detection

    Zhihao Chen, Lei Zhu, Liang Wan, Song Wang, Wei Feng, and Pheng-Ann Heng. A Multi-Task Mean Teacher for Semi-Supervised Shadow Detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , jun 2020

    work page 2020

  8. [7]

    The Pascal Visual Object Classes (VOC) Challenge

    M Everingham, L Van ∼Gool, C K I Williams, J Winn, and A Zisserman. The Pascal Visual Object Classes (VOC) Challenge. ijcv, 88(2):303–338, jun 2010

    work page 2010

Show all 42 references
  1. [8]

    NDDR-CNN: Lay- erwise 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: Lay- erwise Feature Fusing in Multi-Task CNNs by Neural Discriminative Dimensionality Reduction. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition (CVPR), jun 2019

    2019

  2. [9]

    Dynamic Task Prioritization for Multitask Learning

    Michelle Guo, Albert Haque, De-An Huang, Serena Yeung, and Li Fei-Fei. Dynamic Task Prioritization for Multitask Learning. InProceedings of the European Conference on Computer Vision (ECCV), sep 2018

    2018

  3. [10]

    Semantic contours from inverse detectors

    Bharath Hariharan, Pablo Arbeláez, Lubomir Bourdev, Subhransu Maji, and Jitendra Malik. Semantic contours from inverse detectors. In 2011 International Conference on Computer Vision, pages 991–998, 2011

    2011

  4. [11]

    Partly Supervised Multi-Task Learning

    Abdullah-Al-Zubaer Imran, Chao Huang, Hui Tang, Wei Fan, Yuan Xiao, Dingjun Hao, Zhen Qian, and Demetri Terzopoulos. Partly Supervised Multi-Task Learning. In2020 19th IEEE International Conference on Machine Learning and Applications (ICMLA) , pages 769–774, 2020

    2020

  5. [12]

    The distribution of the Flora in the Alpine Zone

    Paul Jaccard. The distribution of the Flora in the Alpine Zone. 1. New Phytologist, 1912. 12 H.-Â. LÊ, P . BERG, M.-T. PHAM: BOMBO: BOX FOR MASK AND MASK FOR BOX

    1912

  6. [13]

    Panoptic Segmentation

    Alexander Kirillov, Kaiming He, Ross Girshick, Carsten Rother, and Piotr Dollar. Panoptic Segmentation. In CVPR, 2019

    2019

  7. [14]

    Efficient inference in fully connected CRFs with Gaussian edge potentials

    Philipp Krähenbühl and Vladlen Koltun. Efficient inference in fully connected CRFs with Gaussian edge potentials. In Proceedings of the 24th International Conference on Neural Information Processing Systems , NIPS’11, pages 109–117, Red Hook, NY , USA, 2011. Curran Associates Inc

    2011

  8. [15]

    Box2Seg: Attention Weighted Loss and Discriminative Feature Learning for Weakly Supervised Segmentation

    Viveka Kulharia, Siddhartha Chandra, Amit Agrawal, Philip Torr, and Ambrish Tyagi. Box2Seg: Attention Weighted Loss and Discriminative Feature Learning for Weakly Supervised Segmentation. In ECCV, 2020

    2020

  9. [17]

    PseCo: Pseudo Labeling and Consistency Training for Semi-Supervised Object Detection

    Gang Li, Xiang Li, Yujie Wang, Shanshan Zhang, Yichao Wu, and Ding Liang. PseCo: Pseudo Labeling and Consistency Training for Semi-Supervised Object Detection. arXiv preprint arXiv:2203.16317, 2022

    2022 arXiv

  10. [18]

    Knowledge Distillation for Multi-task Learning

    Wei-Hong Li and Hakan Bilen. Knowledge Distillation for Multi-task Learning. In eccvw, 2020

    2020

  11. [19]

    Learning Multiple Dense Prediction Tasks from Partially Annotated Data

    Wei-Hong Li, Xialei Liu, and Hakan Bilen. Learning Multiple Dense Prediction Tasks from Partially Annotated Data. In IEEE/CVF International Conference on Computer Vision and Pattern Recognition (CVPR), jun 2022

    2022

  12. [20]

    Microsoft COCO: Common Objects in Context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ra- manan, Piotr Dollar, and C Lawrence Zitnick. Microsoft COCO: Common Objects in Context. In David Fleet, Tomas Pajdla, Bernt Schiele, and Tinne Tuytelaars, editors, eccv, pages 740–755, Cham, 2014....

    2014

  13. [21]

    Feature Pyramid Networks for Object Detection

    Tsung-Yi Lin, Piotr Dollar, Ross Girshick, Kaiming He, Bharath Hariharan, and Serge Belongie. Feature Pyramid Networks for Object Detection. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , jul 2017

    2017

  14. [22]

    Focal Loss for Dense Object Detection

    Tsung-Yi Lin, Priya Goyal, Ross Girshick, Kaiming He, and Piotr Dollar. Focal Loss for Dense Object Detection. In Proceedings of the IEEE International Conference on Computer Vision (ICCV), oct 2017

    2017

  15. [23]

    End-To-End Multi-Task Learning With Attention

    Shikun Liu, Edward Johns, and Andrew J Davison. End-To-End Multi-Task Learning With Attention. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), jun 2019

    2019

  16. [24]

    Path Aggregation Network for Instance Segmentation

    Shu Liu, Lu Qi, Haifang Qin, Jianping Shi, and Jiaya Jia. Path Aggregation Network for Instance Segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), jun 2018

    2018

  17. [25]

    Unbiased Teacher for Semi-Supervised Object Detection

    Yen-Cheng Liu, Chih-Yao Ma, Zijian He, Chia-Wen Kuo, Kan Chen, Peizhao Zhang, Bichen Wu, Zsolt Kira, and Peter Vajda. Unbiased Teacher for Semi-Supervised Object Detection. In ICLR, 2021. H.-Â. LÊ, P . BERG, M.-T. PHAM: BOMBO: BOX FOR MASK AND MASK FOR BOX 13

    2021

  18. [26]

    Unbiased Teacher v2: Semi-Supervised Object Detection for Anchor-Free and Anchor-Based Detectors

    Yen-Cheng Liu, Chih-Yao Ma, and Zsolt Kira. Unbiased Teacher v2: Semi-Supervised Object Detection for Anchor-Free and Anchor-Based Detectors. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 9819–9828, jun 2022

    2022

  19. [27]

    Swin Transformer V2: Scaling Up Capacity and Resolution

    Ze Liu, Han Hu, Yutong Lin, Zhuliang Yao, Zhenda Xie, Yixuan Wei, Jia Ning, Yue Cao, Zheng Zhang, Li Dong, Furu Wei, and Baining Guo. Swin Transformer V2: Scaling Up Capacity and Resolution. In International Conference on Computer Vision and Pattern Recognition (CVPR), 2022

    2022

  20. [28]

    Taskology: Utilizing Task Relations at Scale

    Yao Lu, Soren Pirk, Jan Dlabal, Anthony Brohan, Ankita Pasad, Zhao Chen, Vincent Casser, Anelia Angelova, and Ariel Gordon. Taskology: Utilizing Task Relations at Scale. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 8700–870...

    2021

  21. [29]

    Libra R-CNN: Towards Balanced Learning for Object Detection

    Jiangmiao Pang, Kai Chen, Jianping Shi, Huajun Feng, Wanli Ouyang, and Dahua Lin. Libra R-CNN: Towards Balanced Learning for Object Detection. In IEEE Conference on Computer Vision and Pattern Recognition, 2019

    2019

  22. [30]

    Efficient Parametrization of Multi-Domain Deep Neural Networks

    Sylvestre-Alvise Rebuffi, Hakan Bilen, and Andrea Vedaldi. Efficient Parametrization of Multi-Domain Deep Neural Networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), jun 2018

    2018

  23. [31]

    GrabCut: Interactive Fore- ground Extraction Using Iterated Graph Cuts

    Carsten Rother, Vladimir Kolmogorov, and Andrew Blake. GrabCut: Interactive Fore- ground Extraction Using Iterated Graph Cuts. ACM Transactions on Graphics (SIG- GRAPH), 23(3):309–314, aug 2004. ISSN 0730-0301

    2004

  24. [32]

    Learning To Relate Depth and Semantics for Unsupervised Domain Adaptation

    Suman Saha, Anton Obukhov, Danda Pani Paudel, Menelaos Kanakis, Yuhua Chen, Stamatios Georgoulis, and Luc Van Gool. Learning To Relate Depth and Semantics for Unsupervised Domain Adaptation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (...

    2021

  25. [33]

    Box-Driven Class- Wise Region Masking and Filling Rate Guided Loss for Weakly Supervised Semantic Segmentation

    Chunfeng Song, Yan Huang, Wanli Ouyang, and Liang Wang. Box-Driven Class- Wise Region Masking and Filling Rate Guided Loss for Weakly Supervised Semantic Segmentation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), jun 2019

    2019

  26. [34]

    Weakly Supervised Semantic Segmentation via Box-Driven Masking and Filling Rate Shifting

    Chunfeng Song, Wanli Ouyang, and Zhaoxiang Zhang. Weakly Supervised Semantic Segmentation via Box-Driven Masking and Filling Rate Shifting. TPAMI, (12), 2023

    2023

  27. [35]

    ZipIt! Merging Models from Different Tasks without Training

    George Stoica, Daniel Bolya, Jakob Bjorner, Taylor Hearn, and Judy Hoffman. ZipIt! Merging Models from Different Tasks without Training. International Conference on Learning Representations (ICLR), 2024

    2024

  28. [36]

    Multi-Task Learning for Dense Prediction Tasks: A Survey

    S Vandenhende, S Georgoulis, W Van Gansbeke, M Proesmans, D Dai, and L Van Gool. Multi-Task Learning for Dense Prediction Tasks: A Survey. IEEE Transac- tions on Pattern Analysis and Machine Intelligence, 44(07):3614–3633, jul 2022. ISSN 1939-3539

    2022

  29. [37]

    Semi-supervised Object Detection: A Survey on Recent Research and Progress, 2023

    Yanyang Wang, Zhaoxiang Liu, and Shiguo Lian. Semi-supervised Object Detection: A Survey on Recent Research and Progress, 2023. 14 H.-Â. LÊ, P . BERG, M.-T. PHAM: BOMBO: BOX FOR MASK AND MASK FOR BOX

    2023

  30. [38]

    De- tectron2, 2019

    Yuxin Wu, Alexander Kirillov, Francisco Massa, Wan-Yen Lo, and Ross Girshick. De- tectron2, 2019

    2019

  31. [39]

    End-to-End Semi-Supervised Object Detection With Soft Teacher

    Mengde Xu, Zheng Zhang, Han Hu, Jianfeng Wang, Lijuan Wang, Fangyun Wei, Xi- ang Bai, and Zicheng Liu. End-to-End Semi-Supervised Object Detection With Soft Teacher. In Proceedings of the IEEE/CVF International Conference on Computer Vi- sion (ICCV), pages 3060–3069, oct 2021

    2021

  32. [40]

    Robust Learning Through Cross-Task Consistency

    Amir R Zamir, Alexander Sax, Nikhil Cheerla, Rohan Suri, Zhangjie Cao, Jitendra Malik, and Leonidas J Guibas. Robust Learning Through Cross-Task Consistency. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), jun 2020

    2020

  33. [41]

    Joint Task-Recursive Learning for Semantic Segmentation and Depth Estimation

    Zhenyu Zhang, Zhen Cui, Chunyan Xu, Zequn Jie, Xiang Li, and Jian Yang. Joint Task-Recursive Learning for Semantic Segmentation and Depth Estimation. In Vitto- rio Ferrari, Martial Hebert, Cristian Sminchisescu, and Yair Weiss, editors, Computer Vision – ECCV 2018, pages 238–2...

    2018

  34. [42]

    Instant-Teaching: An End-to-End Semi-Supervised Object Detection Framework

    Qiang Zhou, Chaohui Yu, Zhibin Wang, Qi Qian, and Hao Li. Instant-Teaching: An End-to-End Semi-Supervised Object Detection Framework. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 4081–4090, jun 2021. H.-Â. LÊ, P . BERG, M....

    2021

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.