mlsanityData to 7 Oct 2026

MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer

arXiv 2110.02178 ↗First posted 5 Oct 2021cs.CV cs.AI cs.LGcited by 636

Not on the map. Why?

Abstract (arXiv)

Light-weight convolutional neural networks (CNNs) are the de-facto for mobile vision tasks. Their spatial inductive biases allow them to learn representations with fewer parameters across different vision tasks. However, these networks are spatially local. To learn global representations, self-attention-based vision trans-formers (ViTs) have been adopted. Unlike CNNs, ViTs are heavy-weight. In this paper, we ask the following question: is it possible to combine the strengths of CNNs and ViTs to build a light-weight and low latency network for mobile vision tasks? Towards this end, we introduce MobileViT, a light-weight and general-purpose vision transformer for mobile devices. MobileViT presents a different perspective for the global processing of information with transformers, i.e., transformers as convolutions. Our results show that MobileViT significantly outperforms CNN- and ViT-based networks across different tasks and datasets. On the ImageNet-1k dataset, MobileViT achieves top-1 accuracy of 78.4% with about 6 million parameters, which is 3.2% and 6.2% more accurate than MobileNetv3 (CNN-based) and DeIT (ViT-based) for a similar number of parameters. On the MS-COCO object detection task, MobileViT is 5.7% more accurate than MobileNetv3 for a similar number of parameters. Our source code is open-source and available at: https://github.com/apple/ml-cvnets

Key facts. Cited by 636 papers in the map (data through 7 Oct 2026). Builds on 56 of its 65 references that are in the map. First posted on arXiv 5 Oct 2021.

Builds on

56 of 65 references.

References · 56

  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. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  5. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  6. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,441
  7. 2016Densely Connected Convolutional Networkscited by 6,691
  8. 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,415
  9. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  10. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 4,993
  11. 2014Fully Convolutional Networks for Semantic Segmentationcited by 4,845
  12. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,773
38 more
  1. 2017Squeeze-and-Excitation Networkscited by 4,763
  2. 2020Training data-efficient image transformers & distillation through attentioncited by 4,560
  3. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,827
  4. 2015SSD: Single Shot MultiBox Detectorcited by 3,125
  5. 2016Xception: Deep Learning with Depthwise Separable Convolutionscited by 2,320
  6. 2017Non-local Neural Networkscited by 2,290
  7. 2017Rethinking Atrous Convolution for Semantic Image Segmentationcited by 2,154
  8. 2019Searching for MobileNetV3cited by 1,984
  9. 2021Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutionscited by 1,960
  10. 2019Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelismcited by 1,829
  11. 2021Vision Transformers for Dense Predictioncited by 1,789
  12. 2020GShard: Scaling Giant Models with Conditional Computation and Automatic Shardingcited by 1,601
  13. 2017Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inferencecited by 1,548
  14. 2020Linformer: Self-Attention with Linear Complexitycited by 1,462
  15. 2016YOLO9000: Better, Faster, Strongercited by 1,429
  16. 2017Random Erasing Data Augmentationcited by 1,235
  17. 2018ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Designcited by 1,132
  18. 2015Multi-Scale Context Aggregation by Dilated Convolutionscited by 1,128
  19. 2021Tokens-to-Token ViT: Training Vision Transformers from Scratch on ImageNetcited by 1,096
  20. 2021CvT: Introducing Convolutions to Vision Transformerscited by 946
  21. 2018MnasNet: Platform-Aware Neural Architecture Search for Mobilecited by 937
  22. 2017Sigmoid-Weighted Linear Units for Neural Network Function Approximation in Reinforcement Learningcited by 935
  23. 2021DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsificationcited by 849
  24. 2021CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classificationcited by 806
  25. 2021Going deeper with Image Transformerscited by 759
  26. 2021CoAtNet: Marrying Convolution and Attention for All Data Sizescited by 566
  27. 2019Stand-Alone Self-Attention in Vision Modelscited by 542
  28. 2021ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biasescited by 484
  29. 2021LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inferencecited by 475
  30. 2021Early Convolutions Help Transformers See Bettercited by 395
  31. 2021Bottleneck Transformers for Visual Recognitioncited by 364
  32. 2019Attention Augmented Convolutional Networkscited by 363
  33. 2021Rethinking Spatial Dimensions of Vision Transformerscited by 361
  34. 2021DeepViT: Towards Deeper Vision Transformercited by 288
  35. 2021Incorporating Convolution Designs into Visual Transformerscited by 248
  36. 2021Mobile-Former: Bridging MobileNet and Transformercited by 234
  37. 2021LocalViT: Analyzing Locality in Vision Transformerscited by 228
  38. 2021Multi-Scale Vision Longformer: A New Vision Transformer for High-Resolution Image Encodingcited by 207

50 most cited of 56 All 56 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 5; 2022: 85; 2023: 164; 2024: 150; 2025: 145; 2026: 87.

636 papers.

Sort:

By area:

  1. 2022Token Merging: Your ViT But Fastercited by 995
  2. 2023Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectivescited by 477
  3. 2023Faster Segment Anything: Towards Lightweight SAM for Mobile Applicationscited by 345
  4. 2022EfficientFormer: Vision Transformers at MobileNet Speedcited by 240
  5. 2022TinyViT: Fast Pretraining Distillation for Small Vision Transformerscited by 232
  6. 2022UniFormer: Unifying Convolution and Self-attention for Visual Recognitioncited by 206
  7. 2023EfficientSAM: Leveraged Masked Image Pretraining for Efficient Segment Anythingcited by 187
  8. 2023EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attentioncited by 172
  9. 2023FLatten Transformer: Vision Transformer using Focused Linear Attentioncited by 170
  10. 2022Separable Self-attention for Mobile Vision Transformerscited by 163
  11. 2024EfficientVMamba: Atrous Selective Scan for Light Weight Visual Mambacited by 146
  12. 2022RepQ-ViT: Scale Reparameterization for Post-Training Quantization of Vision Transformerscited by 134
38 more
  1. 2023Run, Don't Walk: Chasing Higher FLOPS for Faster Neural Networkscited by 129
  2. 2024OMG-LLaVA: Bridging Image-level, Object-level, Pixel-level Reasoning and Understandingcited by 126
  3. 2023Transformer-Based Visual Segmentation: A Surveycited by 124
  4. 2023RepViT: Revisiting Mobile CNN From ViT Perspectivecited by 123
  5. 2023Token Fusion: Bridging the Gap between Token Pruning and Token Mergingcited by 123
  6. 2022EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applicationscited by 121
  7. 2022Rethinking Vision Transformers for MobileNet Size and Speedcited by 121
  8. 2022EdgeViTs: Competing Light-weight CNNs on Mobile Devices with Vision Transformerscited by 116
  9. 2022MobileOne: An Improved One millisecond Mobile Backbonecited by 109
  10. 2024MobileNetV4 -- Universal Models for the Mobile Ecosystemcited by 98
  11. 2023MedViT: A Robust Vision Transformer for Generalized Medical Image Classificationcited by 89
  12. 2023CLIP-KD: An Empirical Study of CLIP Model Distillationcited by 87
  13. 2024A Survey of Resource-efficient LLM and Multimodal Foundation Modelscited by 84
  14. 2022TopFormer: Token Pyramid Transformer for Mobile Semantic Segmentationcited by 82
  15. 2022Conv2Former: A Simple Transformer-Style ConvNet for Visual Recognitioncited by 81
  16. 2023Full Stack Optimization of Transformer Inference: a Surveycited by 79
  17. 2022Fast Vision Transformers with HiLo Attentioncited by 78
  18. 2022Deep Model Reassemblycited by 78
  19. 2023Content-based Unrestricted Adversarial Attackcited by 73
  20. 2023Towards Accurate Post-Training Quantization for Vision Transformercited by 70
  21. 2023SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision Applicationscited by 66
  22. 2022Next-ViT: Next Generation Vision Transformer for Efficient Deployment in Realistic Industrial Scenarioscited by 65
  23. 2023Omni Aggregation Networks for Lightweight Image Super-Resolutioncited by 62
  24. 2022Transformers in Remote Sensing: A Surveycited by 61
  25. 2023SparseViT: Revisiting Activation Sparsity for Efficient High-Resolution Vision Transformercited by 60
  26. 2022MogaNet: Multi-order Gated Aggregation Networkcited by 59
  27. 2023Radar-Camera Fusion for Object Detection and Semantic Segmentation in Autonomous Driving: A Comprehensive Reviewcited by 57
  28. 2023Spatially-Adaptive Feature Modulation for Efficient Image Super-Resolutioncited by 56
  29. 2022Hydra Attention: Efficient Attention with Many Headscited by 55
  30. 2024SHViT: Single-Head Vision Transformer with Memory Efficient Macro Designcited by 55
  31. 2023Exploring Lightweight Hierarchical Vision Transformers for Efficient Visual Trackingcited by 53
  32. 2023FedMultimodal: A Benchmark For Multimodal Federated Learningcited by 51
  33. 2023Scale-Aware Modulation Meet Transformercited by 49
  34. 2024reBEN: Refined BigEarthNet Dataset for Remote Sensing Image Analysiscited by 48
  35. 2022MOAT: Alternating Mobile Convolution and Attention Brings Strong Vision Modelscited by 47
  36. 2023Detecting Backdoors During the Inference Stage Based on Corruption Robustness Consistencycited by 45
  37. 2023Adaptive Frequency Filters As Efficient Global Token Mixerscited by 45
  38. 2022ShuffleMixer: An Efficient ConvNet for Image Super-Resolutioncited by 44

50 most cited of 636 All 636 (JSON)