Code Recovery

1. Spatially Adaptive Inference with Stochastic Feature Sampling and Interpolation

1.1. GitHub - zdaxie/SpatiallyAdaptiveInference-Detection: Spatially Adaptive Inference with Stochastic Feature Sampling and Interpolation, ECCV 2020 Oral

1.2. Environment

1.2.1. Python 3.7 （3.7.6） PyTorch == 1.6.0 Torchvision == 0.2.1 CUDA 10.2 Other dependencies

1.2.2. Python 3.7 PyTorch == 1.1.0 Torchvision == 0.3.0 CUDA 9.0 Other dependencies

2. Dynamic Convolutions: Exploiting Spatial Sparsity for Faster Inference

2.1. GitHub - thomasverelst/dynconv: Code for Dynamic Convolutions: Exploiting Spatial Sparsity for Faster Inference (CVPR2020)

3. dataset

3.1. imagenet1k有百万数据，imagenet21k有千万数据， JFT300M有亿级别的 3:43 PM cifar才万级别的，比较是和做最初的尝试

3.2. 各领域常见AI公开数据集汇总 · 附下载地址 | 技术博客 | Graviti 格物钛

3.3. pytorch cifar100 classfication参考：GitHub - weiaicunzai/pytorch-cifar100: Practice on cifar100(ResNet, DenseNet, VGG, GoogleNet, InceptionV3, InceptionV4, Inception-ResNetv2, Xception, Resnet In Resnet, ResNext,ShuffleNet, ShuffleNetv2, MobileNet, MobileNetv2, SqueezeNet, NasNet, Residual Attention Network, SENet, WideResNet) pytorch dataset相关教程（官方，后两个做参考）： 1. https://pytorch.org/tutorials/beginner/basics/data_tutorial.html 2. https://pytorch.org/tutorials/beginner/data_loading_tutorial.html 3. https://pytorch.org/docs/stable/data.html

4. Pytorch Document