https://github.com/jianzhnie/cropandresize_pytorch
CropAndResize in pytorch
https://github.com/jianzhnie/cropandresize_pytorch
cropandresize
Last synced: 6 months ago
JSON representation
CropAndResize in pytorch
- Host: GitHub
- URL: https://github.com/jianzhnie/cropandresize_pytorch
- Owner: jianzhnie
- Created: 2019-12-05T01:50:55.000Z (almost 6 years ago)
- Default Branch: master
- Last Pushed: 2019-12-05T01:55:52.000Z (almost 6 years ago)
- Last Synced: 2025-02-14T11:53:05.621Z (8 months ago)
- Topics: cropandresize
- Language: Python
- Homepage:
- Size: 74.7 MB
- Stars: 1
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# CropAndResize for PyTorch
This implementation is PyTorch version of `crop_and_resize` and supports both forward and backward on CPU and GPU. It's faster than PIL based crop and resize.
## Introduction
The `crop_and_resize` function is ported from [tensorflow](https://www.tensorflow.org/api_docs/python/tf/image/crop_and_resize),
and has the same interface with tensorflow version, except the input feature map
should be in `NCHW` order in PyTorch.
They also have the same output value (error < 1e-5) for both forward and backward as we expected.
**Note:**
Document of `crop_and_resize` can be found [here](https://www.tensorflow.org/api_docs/python/tf/image/crop_and_resize).Notice that `crop_and_resize` use *normalized `(y1, x1, y2, x2)`* as input).
**Warning:**
Currently it only works using the default GPU (index 0)
## Usage
+ Install the package and run test.
```sh
cd CropAndResize.pytorch
## make install
python setup.py install
## run test
sh test.sh
```
+ Use crop_and_resize function for student behavior classification.
```sh
cd body_action_classifier
python demo.py
```
+ When batch_size > 1, the processing is slightly different, please check it in minibatch_demo.py.
```sh
cd body_action_classifier
python minibatch_demo.py
```
## Wiki
The result of `crop_and_resize` may different from original PIL based function, you can find the experimental result and there difference [here](https://wiki.dm-ai.cn/pages/viewpage.action?pageId=71787887#).