mlsanityData to 7 Oct 2026

Eliminate Deviation with Deviation for Data Augmentation and a General Multi-modal Data Learning Method

arXiv 2101.08533 ↗First posted 21 Jan 2021cs.CVcited by 42

Not on the map. Why?

Abstract (arXiv)

One of the challenges of computer vision is that it needs to adapt to color deviations in changeable environments. Therefore, minimizing the adverse effects of color deviation on the prediction is one of the main goals of vision task. Current solutions focus on using generative models to augment training data to enhance the invariance of input variation. However, such methods often introduce new noise, which limits the gain from generated data. To this end, this paper proposes a strategy eliminate deviation with deviation, which is named Random Color Dropout (RCD). Our hypothesis is that if there are color deviation between the query image and the gallery image, the retrieval results of some examples will be better after ignoring the color information. Specifically, this strategy balances the weights between color features and color-independent features in the neural network by dropouting partial color information in the training data, so as to overcome the effect of color devitaion. The proposed RCD can be combined with various existing ReID models without changing the learning strategy, and can be applied to other computer vision fields, such as object detection. Experiments on several ReID baselines and three common large-scale datasets such as Market1501, DukeMTMC, and MSMT17 have verified the effectiveness of this method. Experiments on Cross-domain tests have shown that this strategy is significant eliminating the domain gap. Furthermore, in order to understand the working mechanism of RCD, we analyzed the effectiveness of this strategy from the perspective of classification, which reveals that it may be better to utilize many instead of all of color information in visual tasks with strong domain variations.

Key facts. Cited by 42 papers in the map (data through 7 Oct 2026). Builds on 36 of its 65 references that are in the map. First posted on arXiv 21 Jan 2021.

Builds on

36 of 65 references.

References · 36

  1. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  2. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,690
  3. 2019CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Featurescited by 2,875
  4. 2015FaceNet: A Unified Embedding for Face Recognition and Clusteringcited by 2,833
  5. 2017Improved Regularization of Convolutional Neural Networks with Cutoutcited by 1,623
  6. 2017Random Erasing Data Augmentationcited by 1,235
  7. 2017StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translationcited by 1,036
  8. 2017In Defense of the Triplet Loss for Person Re-Identificationcited by 784
  9. 2020Deep Learning for Person Re-identification: A Survey and Outlookcited by 547
  10. 2017Person Transfer GAN to Bridge Domain Gap for Person Re-Identificationcited by 479
  11. 2019Bag of Tricks and A Strong Baseline for Deep Person Re-identificationcited by 429
  12. 2020Circle Loss: A Unified Perspective of Pair Similarity Optimizationcited by 402
24 more
  1. 2017Unlabeled Samples Generated by GAN Improve the Person Re-identification Baseline in vitrocited by 280
  2. 2017Re-ranking Person Re-identification with k-reciprocal Encodingcited by 260
  3. 2019Joint Discriminative and Generative Learning for Person Re-identificationcited by 211
  4. 2016Person Re-identification: Past, Present and Futurecited by 180
  5. 2017Image-Image Domain Adaptation with Preserved Self-Similarity and Domain-Dissimilarity for Person Re-identificationcited by 168
  6. 2020FastReID: A Pytorch Toolbox for General Instance Re-identificationcited by 164
  7. 2019Invariance Matters: Exemplar Memory for Domain Adaptive Person Re-identificationcited by 133
  8. 2019Relation-Aware Global Attention for Person Re-identificationcited by 124
  9. 2017Improving Person Re-identification by Attribute and Identity Learningcited by 95
  10. 2016A Discriminatively Learned CNN Embedding for Person Re-identificationcited by 78
  11. 2019Ranked List Loss for Deep Metric Learningcited by 77
  12. 2017Camera Style Adaptation for Person Re-identificationcited by 67
  13. 2017Deep Representation Learning with Part Loss for Person Re-Identificationcited by 44
  14. 2018Domain Adaptation through Synthesis for Unsupervised Person Re-identificationcited by 44
  15. 2016End-to-End Comparative Attention Networks for Person Re-identificationcited by 42
  16. 2020Transferable, Controllable, and Inconspicuous Adversarial Attacks on Person Re-identification With Deep Mis-Rankingcited by 42
  17. 2020VehicleNet: Learning Robust Visual Representation for Vehicle Re-identificationcited by 42
  18. 2016Gated Siamese Convolutional Neural Network Architecture for Human Re-Identificationcited by 34
  19. 2016A Siamese Long Short-Term Memory Architecture for Human Re-Identificationcited by 32
  20. 2019Hetero-Center Loss for Cross-Modality Person Re-Identificationcited by 30
  21. 2018SphereReID: Deep Hypersphere Manifold Embedding for Person Re-Identificationcited by 27
  22. 2019An Efficient Framework for Visible-Infrared Cross Modality Person Re-Identificationcited by 8
  23. 2019Adversarial Metric Attack and Defense for Person Re-identificationcited by 7
  24. 2017Towards a Principled Integration of Multi-Camera Re-Identification and Tracking through Optimal Bayes Filterscited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 3; 2024: 14; 2025: 20; 2026: 5.

42 papers.

Sort:

By area:

  1. 2024Cross-Modality Perturbation Synergy Attack for Person Re-identificationcited by 45
  2. 2025Towards Universal Modal Tracking with Online Dense Temporal Token Learningcited by 30
  3. 2025IDEA: Inverted Text with Cooperative Deformable Aggregation for Multi-modal Object Re-Identificationcited by 27
  4. 2024Beyond Augmentation: Empowering Model Robustness under Extreme Capture Environmentscited by 23
  5. 2024Learning Pattern-Specific Experts for Time Series Forecasting Under Patch-level Distribution Shiftcited by 23
  6. 2024Phy124: Fast Physics-Driven 4D Content Generation from a Single Imagecited by 22
  7. 2024Beyond Dropout: Robust Convolutional Neural Networks Based on Local Feature Maskingcited by 19
  8. 2024Adversarial Learning for Neural PDE Solvers with Sparse Datacited by 16
  9. 2024Cross-Modality Attack Boosted by Gradient-Evolutionary Multiform Optimizationcited by 11
  10. 2024Cross-Task Attack: A Self-Supervision Generative Framework Based on Attention Shiftcited by 9
  11. 2025Cross-modal Ship Re-Identification via Optical and SAR Imagery: A Novel Dataset and Methodcited by 8
  12. 2023CLIP-ReIdent: Contrastive Training for Player Re-Identificationcited by 7
30 more
  1. 2024Enhancing person re-identification via Uncertainty Feature Fusion Method and Auto-weighted Measure Combinationcited by 6
  2. 2025Beyond Role-Based Surgical Domain Modeling: Generalizable Re-Identification in the Operating Roomcited by 6
  3. 2023Underwater Acoustic Target Recognition based on Smoothness-inducing Regularization and Spectrogram-based Data Augmentationcited by 5
  4. 2024Pose-dIVE: Pose-Diversified Augmentation with Diffusion Model for Person Re-Identificationcited by 5
  5. 2025StruSR: Structure-Aware Symbolic Regression with Physics-Informed Taylor Guidancecited by 5
  6. 2025A Theory-Inspired Framework for Few-Shot Cross-Modal Sketch Person Re-Identificationcited by 5
  7. 2023Advancing underwater acoustic target recognition via adaptive data pruning and smoothness-inducing regularizationcited by 2
  8. 2025Learning with Dual-level Noisy Correspondence for Multi-modal Entity Alignmentcited by 2
  9. 2025Diffusion-Based Image Editing: An Unforeseen Adversary to Robust Invisible Watermarkscited by 2
  10. 2026Vanishing Watermarks: Diffusion-Based Image Editing Undermines Robust Invisible Watermarkingcited by 2
  11. 2024Adversarial multi-task underwater acoustic target recognition: towards robustness against various influential factorscited by 1
  12. 2025Geometrically Regularized Transfer Learning with On-Manifold and Off-Manifold Perturbationcited by 1
  13. 2025CORE-ReID: Comprehensive Optimization and Refinement through Ensemble fusion in Domain Adaptation for person re-identificationcited by 1
  14. 2026Editing on the Generative Manifold: A Theoretical and Empirical Study of General Diffusion-Based Image Editing Trade-offscited by 1
  15. 2024A Review of Image Retrieval Techniques: Data Augmentation and Adversarial Learning Approachescited by 0
  16. 2024MARs: Multi-view Attention Regularizations for Patch-based Feature Recognition of Space Terraincited by 0
  17. 2024Fusing Physics-Driven Strategies and Cross-Modal Adversarial Learning: Toward Multi-Domain Applicationscited by 0
  18. 2025On the Mechanisms of Adversarial Data Augmentation for Robust and Adaptive Transfer Learningcited by 0
  19. 2025GAMA: Geometry-Aware Manifold Alignment via Structured Adversarial Perturbations for Robust Domain Adaptationcited by 0
  20. 2025Disentangled Geometric Alignment with Adaptive Contrastive Perturbation for Reliable Domain Transfercited by 0
  21. 2025Dynamic Epsilon Scheduling: A Multi-Factor Adaptive Perturbation Budget for Adversarial Trainingcited by 0
  22. 2025Dynamic Modality Scheduling for Multimodal Large Models via Confidence, Uncertainty, and Semantic Consistencycited by 0
  23. 2025FADE: Adversarial Concept Erasure in Flow Modelscited by 0
  24. 2025CORE-ReID V2: Advancing the Domain Adaptation for Object Re-Identification with Optimized Training and Ensemble Fusioncited by 0
  25. 2025Robust Concept Erasure in Diffusion Models: A Theoretical Perspective on Security and Robustnesscited by 0
  26. 2025Robust RGB-T Tracking via Learnable Visual Fourier Prompt Fine-tuning and Modality Fusion Prompt Generationcited by 0
  27. 2025DeRainMamba: A Frequency-Aware State Space Model with Detail Enhancement for Image Derainingcited by 0
  28. 2026UBATrack: Spatio-Temporal State Space Model for General Multi-Modal Trackingcited by 0
  29. 2026When Denoising Becomes Unsigning: Theoretical and Empirical Analysis of Watermark Fragility Under Diffusion-Based Image Editingcited by 0
  30. 2026Editing Away the Evidence: Diffusion-Based Image Manipulation and the Failure Modes of Robust Watermarkingcited by 0