mlsanityData to 7 Oct 2026

Self-Supervised Learning in Multi-Task Graphs through Iterative Consensus Shift

arXiv 2103.14417 ↗First posted 26 Mar 2021cs.LG cs.CVcited by 4

Not on the map. Why?

Abstract (arXiv)

The human ability to synchronize the feedback from all their senses inspired recent works in multi-task and multi-modal learning. While these works rely on expensive supervision, our multi-task graph requires only pseudo-labels from expert models. Every graph node represents a task, and each edge learns between tasks transformations. Once initialized, the graph learns self-supervised, based on a novel consensus shift algorithm that intelligently exploits the agreement between graph pathways to generate new pseudo-labels for the next learning cycle. We demonstrate significant improvement from one unsupervised learning iteration to the next, outperforming related recent methods in extensive multi-task learning experiments on two challenging datasets. Our code is available at https://github.com/bit-ml/cshift.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 36 of its 50 references that are in the map. First posted on arXiv 26 Mar 2021.

Builds on

36 of 50 references.

References · 36

  1. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  2. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 5,986
  3. 2020Unsupervised Learning of Visual Features by Contrasting Cluster Assignmentscited by 2,845
  4. 2018Unsupervised Representation Learning by Predicting Image Rotationscited by 1,586
  5. 2019Deep High-Resolution Representation Learning for Visual Recognitioncited by 1,417
  6. 2017Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semanticscited by 1,345
  7. 2015Learning Transferable Features with Deep Adaptation Networkscited by 1,323
  8. 2019Contrastive Multiview Codingcited by 1,297
  9. 2016Colorful Image Colorizationcited by 1,113
  10. 2017CyCADA: Cycle-Consistent Adversarial Domain Adaptationcited by 1,079
  11. 2019The Replica Dataset: A Digital Replica of Indoor Spacescited by 885
  12. 2015A Large Dataset to Train Convolutional Networks for Disparity, Optical Flow, and Scene Flow Estimationcited by 878
24 more
  1. 2020What Makes for Good Views for Contrastive Learning?cited by 843
  2. 2017Unsupervised Learning of Depth and Ego-Motion from Videocited by 778
  3. 2018Taskonomy: Disentangling Task Transfer Learningcited by 631
  4. 2016Deep Transfer Learning with Joint Adaptation Networkscited by 609
  5. 2014Deep Domain Confusion: Maximizing for Domain Invariancecited by 597
  6. 2020Hypersim: A Photorealistic Synthetic Dataset for Holistic Indoor Scene Understandingcited by 593
  7. 2019End-to-End Learning of Visual Representations from Uncurated Instructional Videoscited by 506
  8. 2016FCNs in the Wild: Pixel-level Adversarial and Constraint-based Adaptationcited by 242
  9. 2019Large Scale Adversarial Representation Learningcited by 238
  10. 2016Split-Brain Autoencoders: Unsupervised Learning by Cross-Channel Predictioncited by 217
  11. 2019Use What You Have: Video Retrieval Using Representations From Collaborative Expertscited by 214
  12. 2019A review of domain adaptation without target labelscited by 154
  13. 2017Mind the Class Weight Bias: Weighted Maximum Mean Discrepancy for Unsupervised Domain Adaptationcited by 151
  14. 2020Robust Learning Through Cross-Task Consistencycited by 105
  15. 2019Dense Extreme Inception Network: Towards a Robust CNN Model for Edge Detectioncited by 104
  16. 2018Learning a Text-Video Embedding from Incomplete and Heterogeneous Datacited by 103
  17. 2020Online Deep Clustering for Unsupervised Representation Learningcited by 99
  18. 2020Evolving Losses for Unsupervised Video Representation Learningcited by 74
  19. 2020Superpixel Segmentation with Fully Convolutional Networkscited by 72
  20. 2018Progressive Ensemble Networks for Zero-Shot Recognitioncited by 23
  21. 2019Learning Across Tasks and Domainscited by 16
  22. 2018Improving Image Clustering With Multiple Pretrained CNN Feature Extractorscited by 12
  23. 2020Unsupervised Learning of Video Representations via Dense Trajectory Clusteringcited by 11
  24. 2020Semi-Supervised Learning for Multi-Task Scene Understanding by Neural Graph Consensuscited by 8

Built on by

Cited by year

Citing papers per year: 2023: 2; 2024: 0; 2025: 2.

4 papers.

  1. 2023Self-supervised Hypergraphs for Learning Multiple World Interpretationscited by 8
  2. 2023Multi-Task Hypergraphs for Semi-supervised Learning using Earth Observationscited by 6
  3. 2025Probabilistic Hyper-Graphs using Multiple Randomly Masked Autoencoders for Semi-supervised Multi-modal Multi-task Learningcited by 1
  4. 2025Multi-modal video data-pipelines for machine learning with minimal human supervisioncited by 0