mlsanityData to 7 Oct 2026

Building Vision Models upon Heat Conduction

arXiv 2405.16555 ↗First posted 26 May 2024cs.CVcited by 7

Not on the map. Why?

Abstract (arXiv)

Visual representation models leveraging attention mechanisms are challenged by significant computational overhead, particularly when pursuing large receptive fields. In this study, we aim to mitigate this challenge by introducing the Heat Conduction Operator (HCO) built upon the physical heat conduction principle. HCO conceptualizes image patches as heat sources and models their correlations through adaptive thermal energy diffusion, enabling robust visual representations. HCO enjoys a computational complexity of O(N^1.5), as it can be implemented using discrete cosine transformation (DCT) operations. HCO is plug-and-play, combining with deep learning backbones produces visual representation models (termed vHeat) with global receptive fields. Experiments across vision tasks demonstrate that, beyond the stronger performance, vHeat achieves up to a 3x throughput, 80% less GPU memory allocation, and 35% fewer computational FLOPs compared to the Swin-Transformer. Code is available at https://github.com/MzeroMiko/vHeat.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 55 of its 81 references that are in the map. First posted on arXiv 26 May 2024.

Builds on

55 of 81 references.

References · 55

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  4. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  5. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  6. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  7. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
  8. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  9. 2020Denoising Diffusion Implicit Modelscited by 9,655
  10. 2016Densely Connected Convolutional Networkscited by 6,691
  11. 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,415
  12. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understandingcited by 6,322
38 more
  1. 2023Mamba: Linear-Time Sequence Modeling with Selective State Spacescited by 4,616
  2. 2020Training data-efficient image transformers & distillation through attentioncited by 4,560
  3. 2014Going Deeper with Convolutionscited by 4,231
  4. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,827
  5. 2022A ConvNet for the 2020scited by 3,545
  6. 2016Xception: Deep Learning with Depthwise Separable Convolutionscited by 2,320
  7. 2021Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutionscited by 1,960
  8. 2021SwinIR: Image Restoration Using Swin Transformercited by 1,869
  9. 2017Deformable Convolutional Networkscited by 1,481
  10. 2020Linformer: Self-Attention with Linear Complexitycited by 1,462
  11. 2019Natural Adversarial Examplescited by 1,401
  12. 2019MMDetection: Open MMLab Detection Toolbox and Benchmarkcited by 1,379
  13. 2016Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoisingcited by 1,352
  14. 2018Unified Perceptual Parsing for Scene Understandingcited by 1,265
  15. 2021Swin Transformer V2: Scaling Up Capacity and Resolutioncited by 1,215
  16. 2015Multi-Scale Context Aggregation by Dilated Convolutionscited by 1,128
  17. 2024Vision Mamba: Efficient Visual Representation Learning with Bidirectional State Space Modelcited by 1,051
  18. 2024VMamba: Visual State Space Modelcited by 955
  19. 2020Designing Network Design Spacescited by 954
  20. 2023RWKV: Reinventing RNNs for the Transformer Eracited by 814
  21. 2021CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classificationcited by 806
  22. 2021PVT v2: Improved Baselines with Pyramid Vision Transformercited by 749
  23. 2023Retentive Network: A Successor to Transformer for Large Language Modelscited by 594
  24. 2021CoAtNet: Marrying Convolution and Attention for All Data Sizescited by 566
  25. 2017Convolutional Sequence to Sequence Learningcited by 550
  26. 2017Understanding the Effective Receptive Field in Deep Convolutional Neural Networkscited by 549
  27. 2021CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped Windowscited by 526
  28. 2020Plug-and-Play Image Restoration with Deep Denoiser Priorcited by 467
  29. 2022On the Parameterization and Initialization of Diagonal State Space Modelscited by 418
  30. 2021Bottleneck Transformers for Visual Recognitioncited by 364
  31. 2021XCiT: Cross-Covariance Image Transformerscited by 335
  32. 2022BEiT v2: Masked Image Modeling with Vector-Quantized Visual Tokenizerscited by 298
  33. 2021Focal Self-attention for Local-Global Interactions in Vision Transformerscited by 286
  34. 2021Global Filter Networks for Image Classificationcited by 257
  35. 2022Neighborhood Attention Transformercited by 235
  36. 2021Scaling Local Self-Attention for Parameter Efficient Visual Backbonescited by 220
  37. 2016Training Deep Spiking Neural Networks using Backpropagationcited by 203
  38. 2022DaViT: Dual Attention Vision Transformerscited by 174

50 most cited of 55 All 55 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 2; 2026: 5.

7 papers.

  1. 2026WaveFormer: Frequency-Time Decoupled Vision Modeling with Wave Equationcited by 4
  2. 2026A Survey on the Green Development of Large Models: From Resource-Efficient Architectures to Hardware-Software Co-Designcited by 3
  3. 2025LSP-ST: Ladder Shape-Biased Side-Tuning for Robust Infrared Small Target Detectioncited by 0
  4. 2025Enhancing Medical Image Segmentation via Heat Conduction Equationcited by 0
  5. 2026Uncertainty-Guided Adverse Weather Restoration via Gated Transformer Networkcited by 0
  6. 2026Toward Interpretable Multimodal Fusion: Heat Conduction Modeling for Hyperspectral and LiDAR Joint Classificationcited by 0
  7. 2026TailProp: content-adaptive light- and heavy-tailed propagation for visioncited by 0