/LIP

Code and pretrained models for LIP: Local Importance-based Pooling (ICCV 19)

Primary LanguagePythonMIT LicenseMIT

LIP: Local Importance-based Pooling

PyTorch implementations of LIP (ICCV 2019).

[arxiv link]

This codebase is now complete and it contains:

  • the implementation of LIP based on PyTorch primitives,
  • LIP-ResNet,
  • LIP-DenseNet,
  • ImageNet training and testing code,
  • CUDA implementation of LIP.

NEWS

[8.13] We uploaded the code of LIP-ResNets and their ImageNet pretrained models (LIP-ResNet-50 & 101).

[8.17] Fixed the missing init_lr key and the possible in_place mul_ operation problem (reported in PyTorch 1.1).

[9.5] CUDA LIP is now available.

[11.6] The LIP-DensNet model is available (sorry for my procrastination). Change the name of ProjectionLIP to SimplifiedLIP for clarification.

Dependencies

  1. Python 3.6
  2. PyTorch 1.0+
  3. tensorboard and tensorboardX

Pretrained Models

You can download ImageNet pretrained models here.

ImageNet

Please refer to imagenet/README.md.

CUDA LIP

Please refer to cuda-lip/README.md.

Misc

If you find our research helpful, please consider citing our paper.

@InProceedings{LIP_2019_ICCV,
author = {Gao, Ziteng and Wang, Limin and Wu, Gangshan},
title = {LIP: Local Importance-Based Pooling},
booktitle = {The IEEE International Conference on Computer Vision (ICCV)},
month = {October},
year = {2019}
}