https://github.com/BigDeviltjj/mxnet-cornernet

Reproduce of CornerNet
https://github.com/BigDeviltjj/mxnet-cornernet

Last synced: 2 months ago
JSON representation

Reproduce of CornerNet

• Host: GitHub
• URL: https://github.com/BigDeviltjj/mxnet-cornernet
• Owner: BigDeviltjj
• Created: 2018-10-19T01:20:04.000Z (about 6 years ago)
• Default Branch: master
• Last Pushed: 2019-02-28T08:42:46.000Z (almost 6 years ago)
• Last Synced: 2024-08-01T22:40:23.099Z (5 months ago)
• Language: Cuda
• Homepage:
• Size: 13.6 MB
• Stars: 22
• Watchers: 3
• Forks: 1
• Open Issues: 3
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• Awesome-MXNet - CornerNet

README

        
## CornerNet

Reproduce of [Cornernet](https://arxiv.org/pdf/1808.01244v1.pdf)

The original pytorch implementation repository is [here](https://github.com/princeton-vl/CornerNet)

## Requirements

* You will need python modules: cv2, matplotlib and numpy.

* To compile corner pooling layer, yo need to install mxnet 1.3.0, then put the files in cxx_operator into src/operator/nn/ in mxnet source code and compile it, then run

```

cd ${YOUR_MXNET_ROOT}

export PYTHONPATH=$(pwd)/lib/libmxnet.so:${PYTHONPATH}

```

to make sure you import the correct mxnet library.

Alternatively, you can uncomment line 92 and 93 and comment line 94, 95 in symbols/cornernet.py  to use python implementation of cornerpooling layer, which would be much slower.

* run init.sh to compile nms and pycocotools

## Demo results

![demo1](https://github.com/BigDeviltjj/mxnet-cornernet/blob/master/images/image_0000.jpg)

![demo2](https://github.com/BigDeviltjj/mxnet-cornernet/blob/master/images/image_0084.jpg)

## mAP

|        Model          | Training data    | Test data |  mAP |

|:-----------------:|:----------------:|:---------:|:----:|

| [CornerNet_coco_511x511](https://drive.google.com/drive/folders/1kPZaK4bRwzVuyij_uC0_niupw-VlLmcV) | train2014+valminusminival2014| minival2014| 38.9|

## TRAIN

You need to put the coco image files in date.

You can change the batch_size in config/cfg.py according to your gpu number and their computation abilies, but make sure that batch_size number is proportional to the number of gpus.

```

python train.py --gpus 0,1

```

## TEST

Download the compressed model from [CornerNet_coco_511x511](https://drive.google.com/drive/folders/1kPZaK4bRwzVuyij_uC0_niupw-VlLmcV) and unzip it then put it in model/, then run

```

python test.py --prefix model/cornernet --epoch 100 --gpus 0

```

if you want to visualize the test results:

```

python test.py --prefix model/cornernet --epoch 100 --gpus 0 --debug True

```

images will be saved in images/