mlsanityData to 14 Oct 2026

Epistemic Graph: A Plug-And-Play Module For Hybrid Representation Learning

arXiv 2305.18731 ↗First posted 30 May 2023cs.CV cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

In recent years, deep models have achieved remarkable success in various vision tasks. However, their performance heavily relies on large training datasets. In contrast, humans exhibit hybrid learning, seamlessly integrating structured knowledge for cross-domain recognition or relying on a smaller amount of data samples for few-shot learning. Motivated by this human-like epistemic process, we aim to extend hybrid learning to computer vision tasks by integrating structured knowledge with data samples for more effective representation learning. Nevertheless, this extension faces significant challenges due to the substantial gap between structured knowledge and deep features learned from data samples, encompassing both dimensions and knowledge granularity. In this paper, a novel Epistemic Graph Layer (EGLayer) is introduced to enable hybrid learning, enhancing the exchange of information between deep features and a structured knowledge graph. Our EGLayer is composed of three major parts, including a local graph module, a query aggregation model, and a novel correlation alignment loss function to emulate human epistemic ability. Serving as a plug-and-play module that can replace the standard linear classifier, EGLayer significantly improves the performance of deep models. Extensive experiments demonstrates that EGLayer can greatly enhance representation learning for the tasks of cross-domain recognition and few-shot learning, and the visualization of knowledge graphs can aid in model interpretation.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 49 of its 70 references that are in the map. First posted on arXiv 30 May 2023.

Builds on

49 of 70 references.

References · 49

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  3. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,361
  4. 2017Graph Attention Networkscited by 8,923
  5. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,498
  6. 2017Inductive Representation Learning on Large Graphscited by 6,320
  7. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,897
  8. 2015Domain-Adversarial Training of Neural Networkscited by 3,044
  9. 2017Prototypical Networks for Few-shot Learningcited by 3,015
  10. 2016Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotationscited by 2,656
  11. 2016Matching Networks for One Shot Learningcited by 2,255
  12. 2017Deep Hashing Network for Unsupervised Domain Adaptationcited by 1,188
37 more
  1. 2017Learning to Compare: Relation Network for Few-Shot Learningcited by 1,143
  2. 2013Spectral Networks and Locally Connected Networks on Graphscited by 1,102
  3. 2019Meta-Learning with Differentiable Convex Optimizationcited by 461
  4. 2018Meta-Learning for Semi-Supervised Few-Shot Classificationcited by 456
  5. 2018TADAM: Task dependent adaptive metric for improved few-shot learningcited by 408
  6. 2016Unsupervised Domain Adaptation with Residual Transfer Networkscited by 341
  7. 2019Multi-Label Image Recognition with Graph Convolutional Networkscited by 322
  8. 2018Meta-Transfer Learning for Few-Shot Learningcited by 306
  9. 2013Zero-Shot Learning Through Cross-Modal Transfercited by 237
  10. 2020Logic Tensor Networkscited by 199
  11. 2018Zero-Shot Object Detectioncited by 189
  12. 2020Meta-Baseline: Exploring Simple Meta-Learning for Few-Shot Learningcited by 179
  13. 2018Zero-shot Recognition via Semantic Embeddings and Knowledge Graphscited by 155
  14. 2018Open Set Domain Adaptation by Backpropagationcited by 150
  15. 2013Zero-Shot Learning by Convex Combination of Semantic Embeddingscited by 135
  16. 2016Harnessing Deep Neural Networks with Logic Rulescited by 135
  17. 2017Semantic Autoencoder for Zero-Shot Learningcited by 133
  18. 2021Learning Graph Embeddings for Compositional Zero-shot Learningcited by 112
  19. 2017Deep Learning for Physical Processes: Incorporating Prior Scientific Knowledgecited by 96
  20. 2018Importance Weighted Adversarial Nets for Partial Domain Adaptationcited by 90
  21. 2018Rethinking Knowledge Graph Propagation for Zero-Shot Learningcited by 87
  22. 2020Boosting Few-Shot Learning With Adaptive Margin Losscited by 83
  23. 2021Exponential Moving Average Normalization for Self-supervised and Semi-supervised Learningcited by 78
  24. 2015Zero-Shot Learning via Semantic Similarity Embeddingcited by 74
  25. 2017Multi-Label Zero-Shot Learning with Structured Knowledge Graphscited by 72
  26. 2020ALP-KD: Attention-Based Layer Projection for Knowledge Distillationcited by 66
  27. 2020Learning to Combine: Knowledge Aggregation for Multi-Source Domain Adaptationcited by 53
  28. 2022Parameter-Efficient and Student-Friendly Knowledge Distillationcited by 47
  29. 2022Rethinking Generalization in Few-Shot Classificationcited by 36
  30. 2016Learning Continuous Semantic Representations of Symbolic Expressionscited by 25
  31. 2021Knowledge Enhanced Machine Learning Pipeline against Diverse Adversarial Attackscited by 24
  32. 2023Supervised Masked Knowledge Distillation for Few-Shot Transformerscited by 23
  33. 2022Self-Promoted Supervision for Few-Shot Transformercited by 22
  34. 2019Knowledge Graph Transfer Network for Few-Shot Recognitioncited by 20
  35. 2021Learning Visual Models using a Knowledge Graph as a Trainercited by 8
  36. 2021Graphical Modeling for Multi-Source Domain Adaptationcited by 7
  37. 2022Learning to Learn Domain-invariant Parameters for Domain Generalizationcited by 3

Built on by

Not cited yet.