Involution: Inverting the Inherence of Convolution for Visual Recognition

Unofficial PyTorch reimplemention of the paper Involution: Inverting the Inherence of Convolution for Visual Recognition by Duo Li, Jie Hu, Changhu Wang et al. published at CVPR 2021.

This repository includes a PyTorch implementation of 2D Involution using C++/OpenMP/CUDA.

Installation

• Default (Use CUDA and OpenMP)

  pip install git+https://github.com/shikishima-TasakiLab/Involution-PyTorch

• No CUDA

  USE_CUDA=0 pip install git+https://github.com/shikishima-TasakiLab/Involution-PyTorch

• No OpenMP

  USE_OPENMP=0 pip install git+https://github.com/shikishima-TasakiLab/Involution-PyTorch

Example Usage

The 2D involution can be used as a nn.Module as follows:

import torch
import torch.nn as nn
from involution import Involution2d

if torch.cuda.is_available():
    device = torch.device("cuda:0")
else:
    device = torch.device("cpu")

inv2d: nn.Module = Involution2d(in_channels=4, out_channels=8).to(device)

x: torch.Tensor = torch.rand(2, 4, 8, 8).to(device)

y: torch.Tensor = inv2d(x)

The 2D involution takes the following parameters:

Parameter	Description	Type	Default
in_channels	Number of input channels.	int	-
out_channels	Number of output channels.	int	-
kernel_size	Kernel size to be used.	int, (int, int)	7
stride	Stride factor to be utilized.	int, (int, int)	1
padding	Padding to be used in unfold operation.	int, (int, int)	3
dilation	Dilation in unfold to be employed.	int, (int, int)	1
groups	Number of groups to be employed.	int	1
bias	If true bias is utilized in each convolution layer.	bool	False
sigma_mapping	Non-linear mapping as introduced in the paper. If none BN + ReLU is utilized.	torch.nn.Module	None
reduce_ratio	Reduce ration of involution channels.	int	1

Reference

@inproceedings{Li2021,
    author = {Li, Duo and Hu, Jie and Wang, Changhu and Li, Xiangtai and She, Qi and Zhu, Lei and Zhang, Tong and Chen, Qifeng},
    title = {Involution: Inverting the Inherence of Convolution for Visual Recognition},
    booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    month = {June},
    year = {2021}
}