mlsanityData to 7 Oct 2026

Many Perception Tasks are Highly Redundant Functions of their Input Data

arXiv 2407.13841 ↗First posted 18 Jul 2024cs.CV cs.LGcited by 4

Not on the map. Why?

Abstract (arXiv)

We show that many perception tasks, from visual recognition, semantic segmentation, optical flow, depth estimation to vocalization discrimination, are highly redundant functions of their input data. Images or spectrograms, projected into different subspaces, formed by orthogonal bases in pixel, Fourier or wavelet domains, can be used to solve these tasks remarkably well regardless of whether it is the top subspace where data varies the most, some intermediate subspace with moderate variability--or the bottom subspace where data varies the least. This phenomenon occurs because different subspaces have a large degree of redundant information relevant to the task.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 37 of its 107 references that are in the map. First posted on arXiv 18 Jul 2024.

Builds on

37 of 107 references.

References · 37

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  3. 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
  4. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,040
  5. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,427
  6. 2016Wide Residual Networkscited by 3,057
  7. 2021Convex optimizationcited by 2,479
  8. 2020Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domainscited by 2,170
  9. 2018On the Spectral Bias of Neural Networkscited by 1,321
  10. 2016Optimization Methods for Large-Scale Machine Learningcited by 1,140
  11. 2018MINE: Mutual Information Neural Estimationcited by 680
  12. 2019Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networkscited by 562
25 more
  1. 2019Benign Overfitting in Linear Regressioncited by 544
  2. 2019Making Convolutional Networks Shift-Invariant Againcited by 390
  3. 2021The Intrinsic Dimension of Images and Its Impact on Learningcited by 307
  4. 2017Emergence of Invariance and Disentanglement in Deep Representationscited by 213
  5. 2018Implicit Self-Regularization in Deep Neural Networks: Evidence from Random Matrix Theory and Implications for Learningcited by 211
  6. 2017Super-Convergence: Very Fast Training of Neural Networks Using Large Learning Ratescited by 196
  7. 2019Towards Understanding the Spectral Bias of Deep Learningcited by 173
  8. 2020Improving KernelSHAP: Practical Shapley Value Estimation via Linear Regressioncited by 153
  9. 2020Learning Diverse and Discriminative Representations via the Principle of Maximal Coding Rate Reductioncited by 141
  10. 2022Rethinking Vision Transformers for MobileNet Size and Speedcited by 121
  11. 2018AudioMNIST: Exploring Explainable Artificial Intelligence for Audio Analysis on a Simple Benchmarkcited by 72
  12. 2024Learning by Reconstruction Produces Uninformative Features For Perceptioncited by 54
  13. 2021Model Zoo: A Growing "Brain" That Learns Continuallycited by 53
  14. 2023FFCV: Accelerating Training by Removing Data Bottleneckscited by 49
  15. 2014Efficient Estimation of Mutual Information for Strongly Dependent Variablescited by 46
  16. 2023Beyond Normal: On the Evaluation of Mutual Information Estimatorscited by 37
  17. 2023Learning Efficient Coding of Natural Images with Maximum Manifold Capacity Representationscited by 31
  18. 2017Diving into the shallows: a computational perspective on large-scale shallow learningcited by 30
  19. 2023The Training Process of Many Deep Networks Explores the Same Low-Dimensional Manifoldcited by 20
  20. 2014Exploration of synergistic and redundant information sharing in static and dynamical Gaussian systemscited by 15
  21. 2021Does the Data Induce Capacity Control in Deep Learning?cited by 13
  22. 2021Principal Components Bias in Over-parameterized Linear Models, and its Manifestation in Deep Neural Networkscited by 10
  23. 2021A Philosophical Understanding of Representation for Neurosciencecited by 6
  24. 2022A picture of the space of typical learnable taskscited by 6
  25. 2022Target alignment in truncated kernel ridge regressioncited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 3; 2026: 1.

4 papers.

  1. 2025LayerLock: Non-collapsing Representation Learning with Progressive Freezingcited by 3
  2. 2025From Linearity to Non-Linearity: How Masked Autoencoders Capture Spatial Correlationscited by 2
  3. 2025An Analytical Characterization of Sloppiness in Neural Networks: Insights from Linear Modelscited by 0
  4. 2026OctoSense: Self-Supervised Learning for Multimodal Robot Perceptioncited by 0