mlsanityData to 14 Oct 2026

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

arXiv 2411.17536 ↗First posted 26 Nov 2024cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 27 of its 40 references that are in the map. First posted on arXiv 26 Nov 2024.

Builds on

27 of 40 references.

References · 27

  1. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  2. 2016Feature Pyramid Networks for Object Detectioncited by 5,036
  3. 2021Swin Transformer V2: Scaling Up Capacity and Resolutioncited by 1,228
  4. 2018Panoptic Segmentationcited by 802
  5. 2018Path Aggregation Network for Instance Segmentationcited by 798
  6. 2018End-to-End Multi-Task Learning with Attentioncited by 534
  7. 2021End-to-End Semi-Supervised Object Detection with Soft Teachercited by 316
  8. 2021Unbiased Teacher for Semi-Supervised Object Detectioncited by 306
  9. 2019Libra R-CNN: Towards Balanced Learning for Object Detectioncited by 255
  10. 2020Just Pick a Sign: Optimizing Deep Multitask Models with Gradient Sign Dropoutcited by 201
  11. 2023ZipIt! Merging Models from Different Tasks without Trainingcited by 179
  12. 2018Efficient parametrization of multi-domain deep neural networkscited by 168
15 more
  1. 2018NDDR-CNN: Layerwise Feature Fusing in Multi-Task CNNs by Neural Discriminative Dimensionality Reductioncited by 127
  2. 2020TIDE: A General Toolbox for Identifying Object Detection Errorscited by 114
  3. 2021Instant-Teaching: An End-to-End Semi-Supervised Object Detection Frameworkcited by 113
  4. 2020Robust Learning Through Cross-Task Consistencycited by 105
  5. 2019Box-driven Class-wise Region Masking and Filling Rate Guided Loss for Weakly Supervised Semantic Segmentationcited by 98
  6. 2022Unbiased Teacher v2: Semi-supervised Object Detection for Anchor-free and Anchor-based Detectorscited by 69
  7. 2020Knowledge Distillation for Multi-task Learningcited by 51
  8. 2020Automated Search for Resource-Efficient Branched Multi-Task Networkscited by 51
  9. 2022PseCo: Pseudo Labeling and Consistency Training for Semi-Supervised Object Detectioncited by 51
  10. 2019Stochastic Filter Groups for Multi-Task CNNs: Learning Specialist and Generalist Convolution Kernelscited by 42
  11. 2021Learning Multiple Dense Prediction Tasks from Partially Annotated Datacited by 42
  12. 2021Learning to Relate Depth and Semantics for Unsupervised Domain Adaptationcited by 32
  13. 2020Taskology: Utilizing Task Relations at Scalecited by 19
  14. 2020Partly Supervised Multitask Learningcited by 11
  15. 2023Data exploitation: multi-task learning of object detection and semantic segmentation on partially annotated datacited by 3

Built on by

Not cited yet.