Github项目推荐 | AutoML与轻量模型列表

awesome-AutoML-and-Lightweight-Models

by guan-yuan

本项目旨在为自动化研究(特别是轻量级模型)提供信息。有兴趣的同学可以进行收藏或者在Github中推荐/提交项目（论文、项目仓库等）。

高质量（最新）AutoML项目和轻量级模型汇总列表，列表包括以下内容：

• 1.神经结构搜索
• 2.轻量级结构
• 3.模型压缩和加速
• 4.超参数优化
• 5.自动化特征工程

Github项目地址：

https://github.com/guan-yuan/awesome-AutoML-and-Lightweight-Models

详细论文内容，请点击阅读原文后点击相关链接访问。

1 神经结构搜索

[论文]

梯度：

• ASAP: Architecture Search, Anneal and Prune | [2019/04]
• Single-Path NAS: Designing Hardware-Efficient ConvNets in less than 4 Hours | [2019/04]
• dstamoulis/single-path-nas | [Tensorflow]
• Automatic Convolutional Neural Architecture Search for Image Classification Under Different Scenes | [IEEE Access 2019]
• sharpDARTS: Faster and More Accurate Differentiable Architecture Search | [2019/03]
• Learning Implicitly Recurrent CNNs Through Parameter Sharing | [ICLR 2019]
• lolemacs/soft-sharing | [Pytorch]
• Probabilistic Neural Architecture Search | [2019/02]
• Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation | [2019/01]
• SNAS: Stochastic Neural Architecture Search | [ICLR 2019]
• FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Search | [2018/12]
• Neural Architecture Optimization | [NIPS 2018]
• renqianluo/NAO | [Tensorflow]
• DARTS: Differentiable Architecture Search | [2018/06]
• quark0/darts | [Pytorch]
• khanrc/pt.darts | [Pytorch]
• dragen1860/DARTS-PyTorch | [Pytorch]

强化学习：

• Template-Based Automatic Search of Compact Semantic Segmentation Architectures | [2019/04]
• Understanding Neural Architecture Search Techniques | [2019/03]
• Fast, Accurate and Lightweight Super-Resolution with Neural Architecture Search | [2019/01]
• falsr/FALSR | [Tensorflow]
• Multi-Objective Reinforced Evolution in Mobile Neural Architecture Search | [2019/01]
• moremnas/MoreMNAS | [Tensorflow]
• ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware | [ICLR 2019]
• MIT-HAN-LAB/ProxylessNAS | [Pytorch, Tensorflow]
• Transfer Learning with Neural AutoML | [NIPS 2018]
• Learning Transferable Architectures for Scalable Image Recognition | [2018/07]
• wandering007/nasnet-pytorch | [Pytorch]
• tensorflow/models/research/slim/nets/nasnet | [Tensorflow]
• MnasNet: Platform-Aware Neural Architecture Search for Mobile | [2018/07]
• AnjieZheng/MnasNet-PyTorch | [Pytorch]
• Practical Block-wise Neural Network Architecture Generation | [CVPR 2018]
• Efficient Neural Architecture Search via Parameter Sharing | [ICML 2018]
• melodyguan/enas | [Tensorflow]
• carpedm20/ENAS-pytorch | [Pytorch]
• Efficient Architecture Search by Network Transformation | [AAAI 2018]

进化算法：

• Single Path One-Shot Neural Architecture Search with Uniform Sampling | [2019/04]
• DetNAS: Neural Architecture Search on Object Detection | [2019/03]
• The Evolved Transformer | [2019/01]
• Designing neural networks through neuroevolution | [Nature Machine Intelligence 2019]
• EAT-NAS: Elastic Architecture Transfer for Accelerating Large-scale Neural Architecture Search | [2019/01]
• Efficient Multi-objective Neural Architecture Search via Lamarckian Evolution | [ICLR 2019]

SMBO（Sequential Model-Based Optimization - 基于序列模型的优化）：

• MFAS: Multimodal Fusion Architecture Search | [CVPR 2019]
• DPP-Net: Device-aware Progressive Search for Pareto-optimal Neural Architectures | [ECCV 2018]
• Progressive Neural Architecture Search | [ECCV 2018]
• titu1994/progressive-neural-architecture-search | [Keras, Tensorflow]
• chenxi116/PNASNet.pytorch | [Pytorch]

随机搜索：

• Exploring Randomly Wired Neural Networks for Image Recognition | [2019/04]
• Searching for Efficient Multi-Scale Architectures for Dense Image Prediction | [NIPS 2018]

超网络：

• Graph HyperNetworks for Neural Architecture Search | [ICLR 2019]

贝叶斯优化：

• Inductive Transfer for Neural Architecture Optimization | [2019/03]

偏序修剪：

• Partial Order Pruning: for Best Speed/Accuracy Trade-off in Neural Architecture Search | [CVPR 2019]
• lixincn2015/Partial-Order-Pruning | [Caffe]

知识提炼：

• Improving Neural Architecture Search Image Classifiers via Ensemble Learning | [2019/03]

[项目]

• Microsoft/nni | [Python]

2 轻量级结构

[论文]

分割：

• ESPNetv2: A Light-weight, Power Efficient, and General Purpose Convolutional Neural Network | [2018/11]
• sacmehta/ESPNetv2 | [Pytorch]
• ESPNet: Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation | [ECCV 2018]
• sacmehta/ESPNet | [Pytorch]
• BiSeNet: Bilateral Segmentation Network for Real-time Semantic Segmentation | [ECCV 2018]
• ooooverflow/BiSeNet | [Pytorch]
• ycszen/TorchSeg | [Pytorch]
• ERFNet: Efficient Residual Factorized ConvNet for Real-time Semantic Segmentation | [T-ITS 2017]
• Eromera/erfnet_pytorch | [Pytorch]

物体检测：

• Pooling Pyramid Network for Object Detection | [2018/09]
• tensorflow/models | [Tensorflow]
• Tiny-DSOD: Lightweight Object Detection for Resource-Restricted Usages | [BMVC 2018]
• lyxok1/Tiny-DSOD | [Caffe]
• Pelee: A Real-Time Object Detection System on Mobile Devices | [NeurIPS 2018]
• Robert-JunWang/Pelee | [Caffe]
• Robert-JunWang/PeleeNet | [Pytorch]
• Receptive Field Block Net for Accurate and Fast Object Detection | [ECCV 2018]
• ruinmessi/RFBNet | [Pytorch]
• ShuangXieIrene/ssds.pytorch | [Pytorch]
• lzx1413/PytorchSSD | [Pytorch]
• FSSD: Feature Fusion Single Shot Multibox Detector | [2017/12]
• ShuangXieIrene/ssds.pytorch | [Pytorch]
• lzx1413/PytorchSSD | [Pytorch]
• dlyldxwl/fssd.pytorch | [Pytorch]
• Feature Pyramid Networks for Object Detection | [CVPR 2017]
• tensorflow/models | [Tensorflow]

3 模型压缩和加速

[论文]

压缩：

• Slimmable Neural Networks | [ICLR 2019]
• JiahuiYu/slimmable_networks | [Pytorch]
• AMC: AutoML for Model Compression and Acceleration on Mobile Devices | [ECCV 2018]
• AutoML for Model Compression (AMC): Trials and Tribulations | [Pytorch]
• Learning Efficient Convolutional Networks through Network Slimming | [ICCV 2017]
• foolwood/pytorch-slimming | [Pytorch]
• Channel Pruning for Accelerating Very Deep Neural Networks | [ICCV 2017]
• yihui-he/channel-pruning | [Caffe]
• Pruning Convolutional Neural Networks for Resource Efficient Inference | [ICLR 2017]
• jacobgil/pytorch-pruning | [Pytorch]
• Pruning Filters for Efficient ConvNets | [ICLR 2017]

加速：

• Fast Algorithms for Convolutional Neural Networks | [CVPR 2016]
• andravin/wincnn | [Python]

[项目]

• NervanaSystems/distiller | [Pytorch]
• Tencent/PocketFlow | [Tensorflow]

[教程/博客]

• Introducing the CVPR 2018 On-Device Visual Intelligence Challenge

4 超参数优化

[论文]

• Tuning Hyperparameters without Grad Students: Scalable and Robust Bayesian Optimisation with Dragonfly | [2019/03]
• dragonfly/dragonfly
• Google vizier: A service for black-box optimization | [SIGKDD 2017]

[项目]

• Microsoft/nni | [Python]
• dragonfly/dragonfly | [Python]

[教程/博客]

• Hyperparameter tuning in Cloud Machine Learning Engine using Bayesian Optimization
• Overview of Bayesian Optimization
• Bayesian optimization
• krasserm/bayesian-machine-learning | [Python]

5 自动化特征工程

[模型分析器]

• Netscope CNN Analyzer | [Caffe]
• sksq96/pytorch-summary | [Pytorch]
• Lyken17/pytorch-OpCounter | [Pytorch]

[参考]

• LITERATURE ON NEURAL ARCHITECTURE SEARCH
• handong1587/handong1587.github.io
• hibayesian/awesome-automl-papers
• mrgloom/awesome-semantic-segmentation
• amusi/awesome-object-detection