mlsanityData to 13 Oct 2026

Sill-Net: Feature Augmentation with Separated Illumination Representation

arXiv 2102.03539 ↗First posted 6 Feb 2021cs.CVcited by 7

Not on the map. Why?

Abstract (arXiv)

For visual object recognition tasks, the illumination variations can cause distinct changes in object appearance and thus confuse the deep neural network based recognition models. Especially for some rare illumination conditions, collecting sufficient training samples could be time-consuming and expensive. To solve this problem, in this paper we propose a novel neural network architecture called Separating-Illumination Network (Sill-Net). Sill-Net learns to separate illumination features from images, and then during training we augment training samples with these separated illumination features in the feature space. Experimental results demonstrate that our approach outperforms current state-of-the-art methods in several object classification benchmarks.

Key facts. Cited by 7 papers in the map (data through 13 Oct 2026). Builds on 55 of its 87 references that are in the map. First posted on arXiv 6 Feb 2021.

Builds on

55 of 87 references.

References · 55

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,950
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,321
  3. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,476
  4. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,170
  5. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,870
  6. 2016Wide Residual Networkscited by 3,063
  7. 2017Prototypical Networks for Few-shot Learningcited by 3,011
  8. 2019RandAugment: Practical automated data augmentation with a reduced search spacecited by 2,266
  9. 2016Matching Networks for One Shot Learningcited by 2,251
  10. 2015Spatial Transformer Networkscited by 1,425
  11. 2017Random Erasing Data Augmentationcited by 1,238
  12. 2016Instance Normalization: The Missing Ingredient for Fast Stylizationcited by 1,193
38 more
  1. 2017Learning to Compare: Relation Network for Few-Shot Learningcited by 1,141
  2. 2017A Closer Look at Memorization in Deep Networkscited by 939
  3. 2019A Closer Look at Few-shot Classificationcited by 683
  4. 2020Simple Copy-Paste is a Strong Data Augmentation Method for Instance Segmentationcited by 596
  5. 2018Meta-learning with differentiable closed-form solverscited by 396
  6. 2018Dynamic Few-Shot Visual Learning without Forgettingcited by 385
  7. 2017The Effectiveness of Data Augmentation in Image Classification using Deep Learningcited by 368
  8. 2017Data Augmentation Generative Adversarial Networkscited by 278
  9. 2017Meta Networkscited by 269
  10. 2020ClassMix: Segmentation-Based Data Augmentation for Semi-Supervised Learningcited by 209
  11. 2019Learning Data Augmentation Strategies for Object Detectioncited by 139
  12. 2020Feature Space Augmentation for Long-Tailed Datacited by 136
  13. 2017Dataset Augmentation in Feature Spacecited by 124
  14. 2017Few-Shot Image Recognition by Predicting Parameters from Activationscited by 122
  15. 2018Robustly Disentangled Causal Mechanisms: Validating Deep Representations for Interventional Robustnesscited by 108
  16. 2018GAN Augmentation: Augmenting Training Data using Generative Adversarial Networkscited by 107
  17. 2017Multi-Level Variational Autoencoder: Learning Disentangled Representations from Grouped Observationscited by 100
  18. 2018Data Augmentation by Pairing Samples for Images Classificationcited by 96
  19. 2019Prototype Rectification for Few-Shot Learningcited by 96
  20. 2016Disentangling factors of variation in deep representations using adversarial trainingcited by 87
  21. 2020On Mixup Regularizationcited by 82
  22. 2020Leveraging the Feature Distribution in Transfer-based Few-Shot Learningcited by 73
  23. 2020Regularizing Deep Networks with Semantic Data Augmentationcited by 73
  24. 2020On Feature Normalization and Data Augmentationcited by 71
  25. 2018ELEGANT: Exchanging Latent Encodings with GAN for Transferring Multiple Face Attributescited by 50
  26. 2018Disentangling Factors of Variation with Cycle-Consistent Variational Auto-Encoderscited by 49
  27. 2020DecAug: Out-of-Distribution Generalization via Decomposed Feature Representation and Semantic Augmentationcited by 48
  28. 2020On the Generalization Effects of Linear Transformations in Data Augmentationcited by 47
  29. 2017Improving Deep Learning using Generic Data Augmentationcited by 36
  30. 2020Learn to Augment: Joint Data Augmentation and Network Optimization for Text Recognitioncited by 36
  31. 2020TAFSSL: Task-Adaptive Feature Sub-Space Learning for few-shot classificationcited by 30
  32. 2017Improving Shadow Suppression for Illumination Robust Face Recognitioncited by 29
  33. 2015Dreaming More Data: Class-dependent Distributions over Diffeomorphisms for Learned Data Augmentationcited by 24
  34. 2017PatchShuffle Regularizationcited by 23
  35. 2019Variational Prototyping-Encoder: One-Shot Learning with Prototypical Imagescited by 22
  36. 2019STaDA: Style Transfer as Data Augmentationcited by 21
  37. 2018On the Importance of Visual Context for Data Augmentation in Scene Understandingcited by 19
  38. 2016Deep Learning Logo Detection with Data Expansion by Synthesising Contextcited by 13

50 most cited of 55 All 55 (JSON)

Built on by

Cited by year

Citing papers per year: 2021: 1; 2022: 3; 2023: 1; 2024: 2.

7 papers.

  1. 2021Squeezing Backbone Feature Distributions to the Max for Efficient Few-Shot Learningcited by 14
  2. 2023Prototypes-oriented Transductive Few-shot Learning with Conditional Transportcited by 8
  3. 2022The Self-Optimal-Transport Feature Transformcited by 7
  4. 2022Weakly Supervised Disentangled Representation for Goal-conditioned Reinforcement Learningcited by 2
  5. 2024The Balanced-Pairwise-Affinities Feature Transformcited by 2
  6. 2024Tiny models from tiny data: Textual and null-text inversion for few-shot distillationcited by 1
  7. 2022Class-Specific Channel Attention for Few-Shot Learningcited by 0