mlsanityData to 7 Oct 2026

MutualNet: Adaptive ConvNet via Mutual Learning from Network Width and Resolution

arXiv 1909.12978 ↗First posted 27 Sep 2019cs.CVcited by 36

Not on the map. Why?

Abstract (arXiv)

We propose the width-resolution mutual learning method (MutualNet) to train a network that is executable at dynamic resource constraints to achieve adaptive accuracy-efficiency trade-offs at runtime. Our method trains a cohort of sub-networks with different widths using different input resolutions to mutually learn multi-scale representations for each sub-network. It achieves consistently better ImageNet top-1 accuracy over the state-of-the-art adaptive network US-Net under different computation constraints, and outperforms the best compound scaled MobileNet in EfficientNet by 1.5%. The superiority of our method is also validated on COCO object detection and instance segmentation as well as transfer learning. Surprisingly, the training strategy of MutualNet can also boost the performance of a single network, which substantially outperforms the powerful AutoAugmentation in both efficiency (GPU search hours: 15000 vs. 0) and accuracy (ImageNet: 77.6% vs. 78.6%). Code is available at urlhttps://github.com/taoyang1122/MutualNet.

Key facts. Cited by 36 papers in the map (data through 7 Oct 2026). First posted on arXiv 27 Sep 2019.

Builds on

No reference list yet.

Built on by

Cited by year

Citing papers per year: 2021: 14; 2022: 14; 2023: 3; 2024: 5.

36 papers.

Sort:

By area:

  1. 2021Split Computing and Early Exiting for Deep Learning Applications: Survey and Research Challengescited by 114
  2. 2022Source-Free Unsupervised Domain Adaptation: A Surveycited by 79
  3. 2021Mutual Contrastive Learning for Visual Representation Learningcited by 42
  4. 2021Dynamic Network Quantization for Efficient Video Inferencecited by 36
  5. 2021Resolution learning in deep convolutional networks using scale-space theorycited by 26
  6. 2024AdaRevD: Adaptive Patch Exiting Reversible Decoder Pushes the Limit of Image Deblurringcited by 25
  7. 2021Multi-Scale Aligned Distillation for Low-Resolution Detectioncited by 24
  8. 2024EventDance: Unsupervised Source-free Cross-modal Adaptation for Event-based Object Recognitioncited by 23
  9. 2023Categories of Response-Based, Feature-Based, and Relation-Based Knowledge Distillationcited by 19
  10. 2021Dynamic-OFA: Runtime DNN Architecture Switching for Performance Scaling on Heterogeneous Embedded Platformscited by 18
  11. 2021Self-Regulation for Semantic Segmentationcited by 18
  12. 2022Adaptive Patch Exiting for Scalable Single Image Super-Resolutioncited by 16
24 more
  1. 2021Degrade is Upgrade: Learning Degradation for Low-light Image Enhancementcited by 13
  2. 2022Super Vision Transformercited by 13
  3. 2021MutualNet: Adaptive ConvNet via Mutual Learning from Different Model Configurationscited by 12
  4. 2021Learning to Disentangle Scenes for Person Re-identificationcited by 10
  5. 2021Single-Training Collaborative Object Detectors Adaptive to Bandwidth and Computationcited by 6
  6. 2021Improved Techniques for Quantizing Deep Networks with Adaptive Bit-Widthscited by 5
  7. 2021EBJR: Energy-Based Joint Reasoning for Adaptive Inferencecited by 4
  8. 2022Med-DANet: Dynamic Architecture Network for Efficient Medical Volumetric Segmentationcited by 4
  9. 2022Slimmable Networks for Contrastive Self-supervised Learningcited by 4
  10. 2022Large-to-small Image Resolution Asymmetry in Deep Metric Learningcited by 4
  11. 2024A Multi-Resolution Mutual Learning Network for Multi-Label ECG Classificationcited by 4
  12. 2022A Closer Look at Branch Classifiers of Multi-exit Architecturescited by 3
  13. 2022FedDCT: Federated Learning of Large Convolutional Neural Networks on Resource Constrained Devices using Divide and Collaborative Trainingcited by 3
  14. 2022RepMode: Learning to Re-parameterize Diverse Experts for Subcellular Structure Predictioncited by 3
  15. 2022Dynamic Low-Resolution Distillation for Cost-Efficient End-to-End Text Spottingcited by 2
  16. 2022R'enyi Divergence Deep Mutual Learningcited by 2
  17. 2023Frame Flexible Networkcited by 2
  18. 2021Prioritized Subnet Sampling for Resource-Adaptive Supernet Trainingcited by 1
  19. 2023Cooperative Learning for Cost-Adaptive Inferencecited by 1
  20. 2024CARIn: Constraint-Aware and Responsive Inference on Heterogeneous Devices for Single- and Multi-DNN Workloadscited by 1
  21. 2024EventDance++: Language-guided Unsupervised Source-free Cross-modal Adaptation for Event-based Object Recognitioncited by 1
  22. 2022DSPNet: Towards Slimmable Pretrained Networks based on Discriminative Self-supervised Learningcited by 0
  23. 2022ClassPruning: Speed Up Image Restoration Networks by Dynamic N:M Pruningcited by 0
  24. 2022TIER-A: Denoising Learning Framework for Information Extractioncited by 0