Real-time network for mobile devices

Google

1. Learning Transferable Architectures for Scalable Image Recognition NASNet
2. MnasNet: Platform-Aware Neural Architecture Search for Mobile MnasNet
3. MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications MobileNets
4. MobileNetV2: Inverted Residuals and Linear Bottlenecks MobileNetV2 MobileNetV2-pytorch
5. MorphNet: Fast & Simple Resource-Constrained Structure Learning of Deep Networks CVPR2018 code
6. Searching for MobileNetV3 MobileNetV3 unofficial implementation MobileNetV3-for-Segmentation official TF Repo
7. Efficient Net: Rethinking Model Scaling for Convolutional Neural Networks. ICML 2019 tfcode unofficial pytorch version
8. MixNet: Mixed Depthwise Convolutional Kernels BMVC2019 Official TF Repo Unofficial pytporch
9. EfficientNetV2: Smaller Models and Faster Training arxiv2021 EfficientNetV2

Microsoft Research

1. Interleaved Group Convolutions for Deep Neural Networks IGCV
2. IGCV2: Interleaved Structured Sparse Convolutional Neural Networks IGCV2
3. IGCV3: Interleaved Low-Rank Group Convolutions for Efficient Deep Neural Networks IGCV3
4. Accelerating Deep Neural Networks with Spatial Bottleneck Modules arxiv2018
5. Dynamic Convolution: Attention over Convolution Kernels arxiv2019

Facebook

1. Drop an Octave: Reducing Spatial Redundancy in Convolutional Neural Networks with Octave Convolution arxiv2019 unofficial implementation OctaveConv_pytorch OctaveConv_MXNet
2. FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Search arxiv2018 code
3. ChamNet: Towards Efficient Network Design through Platform-Aware Model Adaptation arxiv2018 code
4. Hybrid Composition with IdleBlock: More Efficient Networks for Image Recognition arxiv2019
5. FBNetV2: Differentiable Neural Architecture Search for Spatial and Channel Dimensions CVPR2020 FBNetv2
6. FBNetV3: Joint Architecture-Recipe Search using Neural Acquisition Function arxiv2020
7. FP-NAS: Fast Probabilistic Neural Architecture Search arxiv2020
8. Fast and Accurate Model Scaling arxiv2021

Megvill

1. ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices ShuffleNet
2. ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design ShuffleNet V2 Shufflenet-v2-Pytorch
3. ShuffleNetV2+：paper unrelease code ShuffleNet Series by Megvii Research
4. WeightNet: Revisiting the Design Space of Weight Networks ECCV2020 WeightNet
5. RepVGG: Making VGG-style ConvNets Great Again arxiv2020 RepVGG Tsinghua University, MEGVII Technology, etc

Others

1. CondenseNet: An Efficient DenseNet using Learned Group Convolutions CondenseNet
2. CondenseNet V2: Sparse Feature Reactivation for Deep Networks CondenseNetV2
3. ANTNets: Mobile Convolutional Neural Networks for Resource Efficient Image Classification arxiv2019
4. Seesaw-Net: Convolution Neural Network With Uneven Group Convolution arxiv2019
5. ISBNet: Instance-aware Selective Branching Network arxiv2019
6. Multinomial Distribution Learning for Effective Neural Architecture Search arxiv2019 code
7. HGC: Hierarchical Group Convolution for Highcdly Efficient Neural Network arxiv2019
8. DiCENet: Dimension-wise Convolutions for Efficient Networks arxiv2019 code
9. Densely Connected Search Space for More Flexible Neural Architecture Search arxiv2019 code
10. VarGNet: Variable Group Convolutional Neural Network for Efficient Embedded Computing arxiv2019 Horizon Robotics
11. VarGFaceNet: An Efficient Variable Group Convolutional Neural Network for Lightweight Face Recognition VarGFaceNet code Horizon Robotics (1st place in The Lightweight Face Recognition Challenge & Workshop ICCV 2019)
12. Once for All: Train One Network and Specialize it for Efficient Deployment ICLR 2020 ofaNet MIT && MIT-IBM Watson AI Lab
13. LPRNet: Lightweight Deep Network by Low-rank Pointwise Residual Convolution arxiv2019
14. LeanConvNets: Low-cost Yet Effective Convolutional Neural Networks arxiv2019 Ben-Gurion University&&Emory Universit
15. DeepSquare: Boosting the Learning Power of Deep Convolutional Neural Networks with Elementwise Square Operators arxiv2019 ByteDance AI Lab
16. ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks arxiv2019 code Tianjin University
17. XSepConv: Extremely Separated Convolution arxiv2020 Tsinghua University &&University College London
18. GhostNet: More Features from Cheap Operations CVPR2020 TF Repo Huawei Noah’s Ark Lab&&Peking University&&Sydney University
19. Resolution Adaptive Networks for Efficient Inference CVPR2020 RANet Tsinghua && HIT &&SenseTime
20. TResNet: High Performance GPU-Dedicated Architecture arxiv2020 TResNet DAMO Academy, Alibaba Group
21. Neural Architecture Design for GPU-Efficient Networks arxiv2020 Alibaba Group
22. ReXNet: Diminishing Representational Bottleneck on Convolutional Neural Network arxiv2020 rexnet Clova AI Research, NAVER Corp
23. Rethinking Bottleneck Structure for Efficient Mobile Network Design ECCV2020 MobileNeXt yitu-opensource
24. Model Rubik’s Cube: Twisting Resolution, Depth and Width for TinyNets NeurIPS 2020**
25. MicroNet: Towards Image Recognition with Extremely Low FLOPs arxiv2020 UC San Diego && Microsoft
26. Lite-HRNet: A Lightweight High-Resolution Network cvpr2021 Lite-HRNet
27. AsymmNet: Towards ultralight convolution neural networks using asymmetrical bottlenecks MAI@CVPR 2021 AsymmNet

Semantic segmentation

1. ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation ENet
2. ICNet for Real-Time Semantic Segmentation on High-Resolution Images ICNet
3. Speeding up Semantic Segmentation for Autonomous Driving SQNet
4. ERFNet: Efficient Residual Factorized ConvNet for Real-time Semantic Segmentation ERFNet
5. ESPNet: Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation ESPNet
6. BiSeNet: Bilateral Segmentation Network for Real-time Semantic Segmentation ECCV2018 code
7. A Comparative Study of Real-time Semantic Segmentation for Autonomous Driving
8. Efficient Dense Modules of Asymmetric Convolution for Real-Time Semantic Segmentation EDANet
9. Light-Weight RefineNet for Real-Time Semantic Segmentation Light-Weight RefineNet
10. Searching for Efficient Multi-Scale Architectures for Dense Image Prediction
11. CGNet: A Light-weight Context Guided Network for Semantic Segmentation arxiv2018 code
12. ShelfNet for Real-time Semantic Segmentation arxiv2018 code
13. Concentrated-Comprehensive Convolutions for lightweight semantic segmentation arxiv2018
14. ESPNetv2: A Light-weight, Power Efficient, and General Purpose Convolutional Neural Network cvpr2019 code
15. Fast-SCNN: Fast Semantic Segmentation Network arxiv2019 code blog
16. An efficient solution for semantic segmentation: ShuffleNet V2 with atrous separable convolutions arxiv2019
17. Decoders Matter for Semantic Segmentation: Data-Dependent Decoding Enables Flexible Feature Aggregation cvpr2019
18. DFANet: Deep Feature Aggregation for Real-Time Semantic Segmentation cvpr2019Megvii
19. Real time backbone for semantic segmentation arxiv2019
20. Residual Pyramid Learning for Single-Shot Semantic Segmentation arxiv2019
21. Knowledge Adaptation for Efficient Semantic Segmentation cvpr2019
22. In Defense of Pre-trained ImageNet Architectures for Real-time Semantic Segmentation of Road-driving Images cvpr2019 code
23. Towards Real-Time Automatic Portrait Matting on Mobile Devices arxiv2019 code
24. PortraitNet: Real-time Portrait Segmentation Network for Mobile Device Computers & Graphics 2019 code
25. Design of Real-time Semantic Segmentation Decoder for Automated Driving VISAPP2019
26. ThunderNet: A Turbo Unified Network for Real-Time Semantic Segmentation WACV2019
27. LEDNet: A Lightweight Encoder-Decoder Network for Real-time Semantic Segmentation ICIP2019 code
28. Accurate Facial Image Parsing at Real-Time Speed TIP2019
29. Efficient Ladder-style DenseNets for Semantic Segmentation of Large Images arxiv2019
30. Nail Polish Try-On: Realtime Semantic Segmentation of Small Objects forNative and Browser Smartphone AR Applications CVPRW2019
31. ESNet: An Efficient Symmetric Network for Real-time Semantic Segmentation arxiv2019 code
32. Real-time Hair Segmentation and Recoloring on Mobile GPUs 2019CVPRW
33. Efficient Segmentation: Learning Downsampling Near Semantic Boundaries arxiv2019
34. Multi-Class Lane Semantic Segmentation using Efficient Convolutional Networks MMSP 2019
35. A Comparative Study of High-Recall Real-Time Semantic Segmentation Based on Swift Factorized Network SPIE Security + Defence 2019 code
36. Context-Integrated and Feature-Refined Network for Lightweight Urban Scene Parsing arxiv2019
37. DABNet: Depth-wise Asymmetric Bottleneck for Real-time Semantic Segmentation BMVC2019 code
38. Learning Lightweight Lane Detection CNNs by Self Attention Distillation ICCV2019 torchRepo
39. SqueezeNAS: Fast neural architecture search for faster semantic segmentation arxiv2019 code submitted to ICCV Neural Architects workshop
40. ExtremeC3Net: Extreme Lightweight Portrait Segmentation Networks using Advanced C3-modules arxiv2019 ExtremeC3Net Seoul National University && Clova AI, NAVER Corp
41. Customizable Architecture Search for Semantic Segmentation cvpr2019 University of Science and Technology of China && JD AI Research
42. See More than Once – Kernel-Sharing Atrous Convolution for Semantic Segmentation arxiv2019 University of Technology Sydney && East China Normal University
43. Feature Pyramid Encoding Network for Real-time Semantic Segmentation BMVC2019 The University of Manchester
44. Graph-guided Architecture Search for Real-time Semantic Segmentation arxiv2019 SenseTime Research && Zhejiang University
45. Eye Semantic Segmentation with a Lightweight Model ICCVW 2019 code Chonnam National University
46. FDDWNet: A Lightweight Convolutional Neural Network for Real-time Sementic Segmentation arxiv2019 Nanjing University of Posts & Telecommunications
47. Real-Time Semantic Segmentation via Multiply Spatial Fusion Network arxiv2019 Beihang University &&Megvii Technology&&Peng Cheng Laboratory
48. RGPNet: A Real-Time General Purpose Semantic Segmentation arxiv2019 Advanced Research Lab, NavInfo Europe
49. LiteSeg: A Novel Lightweight ConvNet for Semantic Segmentation DICTA2019
50. FasterSeg: Searching for Faster Real-time Semantic Segmentation ICIR2020 FasterSeg Texas A&M University && Horizon Robotics Inc
51. Real-time Segmentation and Facial Skin Tones Grading arxiv2019 HLRNet
52. SINet: Extreme Lightweight Portrait Segmentation Networks with Spatial Squeeze Modules and Information Blocking Decoder WACV2020 SINet
53. FarSee-Net: Real-Time Semantic Segmentation by Efficient Multi-scale Context Aggregation and Feature Space Super-resolution ICRA2020 SenseTime && Tokyo University
54. Cars Can’t Fly up in the Sky: Improving Urban-Scene Segmentation via Height-driven Attention Networks CVPR2020 HANet Korea University && CA
55. Real-Time High-Performance Semantic Image Segmentation of Urban Street Scenes arxiv2020 XMU
56. Real-Time Semantic Segmentation via Auto Depth, Downsampling Joint Decision and Feature Aggregation arxiv2020 ZheJiang University
57. BiSeNet V2: Bilateral Network with Guided Aggregation for Real-time Semantic Segmentation arxiv2020 BiSeNet V2 bisenetv2-tensorflow HUST && The University of Adelaide &&CUHK && Tencent
58. Bi-direction Context Propagation Network for Real-time Semantic Segmentation arxiv2020 Hefei University of Technology
59. LRNNet: A Light-Weighted Network with Efficient Reduced Non-Local Operation for Real-Time Semantic Segmentation arxiv2020 SJTU
60. Real-time Semantic Segmentation with Fast Attention arxiv2020 Homepage FANet Boston University && Adobe Research
61. Ultra Fast Structure-aware Deep Lane Detection ECCV2020 code Zhejiang University
62. Improving Semantic Segmentation via Decoupled Body and Edge Supervision ECCV2020 DecoupleSegNets Peking University && University of Oxford && SenseTime Research && Zhejiang Lab
63. Semantic Segmentation With Multi Scale Spatial Attention For Self Driving Cars arxiv2020
64. EfficientSeg: An Efficient Semantic Segmentation Network arxiv2020 EfficientSeg
65. Dense Dual-Path Network for Real-time Semantic Segmentation ACCV2020
66. CABiNet: Efficient Context Aggregation Network for Low-Latency Semantic Segmentation arxiv2020
67. SegBlocks: Block-Based Dynamic Resolution Networks for Real-Time Segmentation arxiv2020 KU Leuven
68. Real-Time High-Resolution Background Matting arxiv2020 hompage BackgroundMattingV2 University of Washington
69. Is a Green Screen Really Necessary for Real-Time Portrait Matting? arxiv2020 MODNet City University of Hong Kong, SenseTime Research
70. HyperSeg: Patch-wise Hypernetwork for Real-time Semantic Segmentation arxiv2020 Facebook AI, Tel Aviv University
71. Boundary-Aware Segmentation Network for Mobile and Web Applications submitted to TPAMI BASNet
72. EADNet: Efficient Asymmetric Dilated Network for Semantic Segmentation arxiv2021 Fudan University , etc.
73. AttaNet: Attention-Augmented Network for Fast and Accurate Scene Parsing AAAI2021
74. Deep Dual-resolution Networks for Real-time and Accurate Semantic Segmentation of Road Scenes arxiv2021
75. Rethinking BiSeNet For Real-time Semantic Segmentation CVPR2021 STDC-Seg

Instance Segmentation

1. YOLACT: Real-time Instance Segmentation ICCV2019 code University of California, Davis
2. CenterMask:Real-Time Anchor-Free Instance Segmentation VovnetV2 CenterMask CenterMask2 ETRI
3. YOLACT++ Better Real-time Instance Segmentation [arxiv2019](YOLACT++ Better Real-time Instance Segmentation) University of California, Davis
4. BlendMask: Top-Down Meets Bottom-Up for Instance Segmentation arxiv2020 The University of Adelaide && Southeast University &&Huawei Noah’s Ark Lab
5. Deep Snake for Real-Time Instance Segmentation cvpr2020 snake Zhejiang University
6. PolarMask: Single Shot Instance Segmentation with Polar Representation cvpr2020 PolarMask HKU
7. SOLOv2: Dynamic, Faster and Stronger arxiv2020 SOLOV2 UnofficialPytorchRepo The University of Adelaide && Tongji University && ByteDance AI Lab
8. YolactEdge: Real-time Instance Segmentation on the Edge arxiv2020 YolactEdge

Salient Object Detection

• U2-Net: Going Deeper with Nested U-Structure for Salient Object Detection PR2020 U-2-Net
• Highly Efficient Salient Object Detection with 100K Parameters ECCV2020 CSNet
• MobileSal: Extremely Efficient RGB-D Salient Object Detection arxiv2020

Object detection

1. An Energy and GPU-Computation Efficient Backbone Network for Real-Time Object Detection CVPR2019 CEFRL Workshop VOVNetPytorch ETRI
2. EfficientDet: Scalable and Efficient Object Detection arxiv2019 unofficial-EfficientDet.Pytorch Google Research, Brain Team
3. RefineDetLite: A Lightweight One-stage Object Detection Framework for CPU-only Devices arxiv2019 Tencent Research&&HKUST
4. Learning Spatial Fusion for Single-Shot Object Detection arxiv2019 yolov3+ Beihang University
5. CSPNet: A New Backbone that can Enhance Learning Capability of CNN arxiv2019 CSPNet research teams from TaiWan
6. YOLOv4: Optimal Speed and Accuracy of Object Detection arxiv2020 OfficialRepo Minimal PyTorch Tensorflow 2.0Repo KerasRepo
7. YOLO Nano: a Highly Compact You Only Look Once Convolutional Neural Network for Object Detection arxiv2019

Neural Architecture Search

1. PC-DARTS: Partial Channel Connections for Memory-Efficient Differentiable Architecture Search arxiv2019 code Shanghai Jiao Tong University&&Huawei
2. Densely Connected Search Space for More Flexible Neural Architecture Search arxiv2019 code Huazhong University of Science and Technology &&Horizon Robotics
3. FairNAS: Rethinking Evaluation Fairness of Weight Sharing Neural Architecture Search arxiv2019 code Xiaomi AI Lab
4. XferNAS: Transfer Neural Architecture Search arxiv2019 IBM Research
5. AutoML: A Survey of the State-of-the-Art arxiv2019 Hong Kong Baptist University
6. MoGA: Searching Beyond MobileNetV3 arxiv2019 code Xiaomi AI Lab
7. ScarletNAS: Bridging the Gap Between Scalability and Fairness in Neural Architecture Search arxiv2019 code Xiaomi AI Lab && IoT
8. BANANAS: Bayesian Optimization with Neural Architectures for Neural Architecture Search arxiv code blog
9. Fast and Practical Neural Architecture Search iccv2019 CUHK && YouTu Lab, Tencent
10. Fair DARTS: Eliminating Unfair Advantages in Differentiable Architecture Search arxiv2019 FairDARTS Xiaomi AI Lab &&Minzu University of China
11. SGAS: Sequential Greedy Architecture Search arxiv2019 project KAUST && Intel Labs
12. Blockwisely Supervised Neural Architecture Search with Knowledge Distillation arxiv2019 DNA DarkMatter AI Research && Monash University &&Sun Yat-sen University
13. EDAS: Efficient and Differentiable Architecture Search arxiv2019 KAIST
14. Efficient Differentiable Neural Architecture Search with Meta Kernels arxiv2019 HUST &&YITU &&NUS
15. AtomNAS: Fine-Grained End-to-End Neural Architecture Search ICIR2020 AutoNAS Johns Hopkins University && ByteDance AI Lab
16. EcoNAS: Finding Proxies for Economical Neural Architecture Search arxiv2020 The University of Sydney &&Nanyang Technological University&&SenseTime Research
17. MixPath: A Unified Approach for One-shot Neural Architecture Search arxiv2020 MixPath Xiaomi AI Lab &&UCAS

Transformer

1. Efficient Transformers: A Survey arxiv2020
2. Escaping the Big Data Paradigm with Compact Transformers arxiv2021 compact- transformer

Tricks for training for Transformer

1. Training data-efficient image transformers & distillation through attention arxiv2021 DEIT
2. Token Labeling: Training a 85.5% Top-1 Accuracy Vision Transformer with 56M Parameters on ImageNet arxiv2021 TokenLabeling
3. Improve Vision Transformers Training by Suppressing Over-smoothing arxiv2021 PatchVisionTransformer

Tricks for training neural network

1. A Recipe for Training Neural Networks (Apr 25, 2019)
2. Improving deep learning models with bag of tricks (Dec 13,2018)
3. A Bag of Tricks for Image Classification (Dec 17, 2018)
4. Bag of Tricks for Image Classification with Convolutional Neural Networks cvpr2019 code
5. Bag of Freebies for Training Object Detection Neural Networks arxiv2019 code
6. Bag of Tricks for Image Classification by Arthur Kuzin 2020slide
7. carrier-of-tricks-for-classification-pytorch 2020code
8. Faster Deep Learning Training with PyTorch – a 2021 Guide

Some useful links

1. Awesome-model-compression-and-acceleration
2. awesome-model-compression-and-acceleration
3. Model-Compression-Papers
4. Awesome-model-compression-and-acceleration
5. awesome-AutoML-and-Lightweight-Models
6. 常用的语义分割架构结构综述以及代码复现
7. Efficient-Segmentation-Networks

Some blogs

1. Real-time Portrait Segmentation on Smartphones
2. Mobile Real-time Video Segmentation
3. Real-Time deep learning in mobile application
4. QNNPACK: Open source library for optimized mobile deep learning
5. 第十七章 模型压缩及移动端部署
6. Tips for building fast portrait segmentation network with TensorFlow Lite
7. $\color{red}{*}$ Literature On Neural network architecture
8. A 2019 Guide to Semantic Segmentation
9. Semantic Segmentation
10. Image segmentation in 2020: Architectures, Losses, Datasets, and Frameworks
11. New mobile neural network architectures
12. NanoDet

Some projects

1. Real-Time Video Segmentation on Mobile Devices
2. LightNet++: Boosted Light-weighted Networks for Real-time Semantic Segmentation
3. mobile_phone_human_matting
4. Real-Time Semantic Segmentation in Mobile device
5. Semantic Segmentation on PyTorch
6. Lightweight-Segmentation
7. A PyTorch-Based Framework for Deep Learning in Computer Vision
8. A high performance semantic segmentation toolkit based on PaddlePaddle
9. PyTorch image models, scripts, pretrained weights
10. fast semantic segmentation models on CityScapes/Camvid DataSet by Pytorch
11. A semantic segmentation framework by pyotrch
12. PyTorch for Semantic Segmentation

Inference framework

1. ncnn
2. paddle
3. bolt
4. Mace
5. Mnn