https://github.com/escorciav/roi_pooling
ROIPooling for pytorch
https://github.com/escorciav/roi_pooling
conda-environment cupy python37 pytorch roipooling
Last synced: 4 months ago
JSON representation
ROIPooling for pytorch
- Host: GitHub
- URL: https://github.com/escorciav/roi_pooling
- Owner: escorciav
- Created: 2018-02-22T07:02:00.000Z (over 7 years ago)
- Default Branch: master
- Last Pushed: 2019-04-08T12:36:31.000Z (over 6 years ago)
- Last Synced: 2025-06-13T06:07:29.015Z (4 months ago)
- Topics: conda-environment, cupy, python37, pytorch, roipooling
- Language: Python
- Size: 12.7 KB
- Stars: 51
- Watchers: 4
- Forks: 12
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Pytorch ROIPooling
Welcome!
This is a generic implementation of ROIpooling operation used in the context of object detection.
## Feats
- Modularized
- JIT compilation with cupy
- Works well with batches of images :wink:
## Getting started
We need the following requirements `cuda`, `pytorch==1.0.1`, `cupy=5.1.0` which we can get most of them from [anaconda.org](http://anaconda.org/) with trusted channels.
1. Install anaconda or miniconda.
> Skip this if you already have miniconda or anaconda installed in your system.
1. Create a new environment
`conda create -n pytorch-extensions python=3.7 pytorch cupy -c pytorch`
> This step creates a conda environment called `pytorch-extensions`. In case, you change the name keep it mind to update the next lines accordingly.
1. `conda activate pytorch-extensions`
1. `python example.py`
Hopefully everything runs like the breeze.
### Can I use it in Colab?
Sure, take a look at this [notebook](https://colab.research.google.com/drive/1zoZKhWynAwnXJAWGTfOiU2-kbc4FH2EM). It provides a guide for the setup and usage of the `roi_pooling` `Function`.
## LICENSE
[MIT](https://choosealicense.com/licenses/mit/)
We highly appreciate that you leave attribution notes when you copy portions of this codebase in yours.
### Did you like it?
Support me, gimme a :star: in the github banner or invite me a :coffee:/:beer:. If you are in academia, I would appreciate that you cite my research:
```
@article{EscorciaDJGS18,
author = {Victor Escorcia and
Cuong Duc Dao and
Mihir Jain and
Bernard Ghanem and
Cees Snoek},
title = {Guess Where? Actor-Supervision for Spatiotemporal Action Localization},
journal = {CoRR},
volume = {abs/1804.01824},
year = {2018},
url = {http://arxiv.org/abs/1804.01824},
archivePrefix = {arXiv},
eprint = {1804.01824}
}
```
This implementation was built on top of the legendary Faster-RCNN which you _must_ cite:
```
@article{RenHG017,
author = {Shaoqing Ren and
Kaiming He and
Ross B. Girshick and
Jian Sun},
title = {Faster {R-CNN:} Towards Real-Time Object Detection with Region Proposal
Networks},
journal = {{IEEE} Trans. Pattern Anal. Mach. Intell.},
volume = {39},
number = {6},
pages = {1137--1149},
year = {2017},
url = {https://doi.org/10.1109/TPAMI.2016.2577031},
doi = {10.1109/TPAMI.2016.2577031}
}
```
This was also possible due to [Chainer](https://chainer.org/), and the easy to follow [pyinn](https://github.com/szagoruyko/pyinn).
## FAQs
### Do I need to buy an anaconda license?
Of course not! You do everything with virtual environments. Indeed, I would be pleased to accept a PR with a recipe for virtual environments.
_Why anaconda?_
In short, due to the last five letters.
### Why another ROIpooling operation?
Well, I tried many C extensions mainly taken from this [repo](https://github.com/longcw/faster_rcnn_pytorch) but those did not fit my purpose of ROIPooling over batches of images.
_Why?_
You can clearly see [here](https://github.com/longcw/faster_rcnn_pytorch/blob/master/faster_rcnn/roi_pooling/src/roi_pooling_cuda.c#L27-L30) that when the batch size is greater than 1, the output is zero.
_Does that mean that they are useless?_
Of course not! I noticed that FastRCNN uses a batch size of 1. Probably, they did not mind to make it more general implementation.
_Why didn't you remove the conditional?_
I tried in one of the repos but it fails. I even removed all the binaries and compiled again but it still returned zeros. Thus, I just moved on and pursue my personal reason:
I was really curious of launching cupy kernels using data from pytorch tensors. It is simply amazing. Moreover, it was a great experience to explore CUDA and pytorch.autograd.