CenterNet: Keypoint Triplets for Object Detection

by Kaiwen Duan, Song Bai, Lingxi Xie, Honggang Qi, Qingming Huang and Qi Tian

The code to train and eval our CenterNet is available here. Thanks Princeton Vision & Learning Lab! For more details, please refer to our arXiv paper.

Our method is an one-stage detector and learns from scratch. On the MS-COCO dataset, CenterNet achieves an AP of 47.0%！

Introduction

CenterNet is a framework for object detection with deep ConvNets. You can use the code to train/evaluate a network for object detection task.

It achieves state-of-the-art performance on one of the most challenging dataset: MS-COCO.
Our code is written by Python, based on CornerNet.

Architecture

Comparison with other methods

In terms of speed, we test the inference speed of both CornerNet and CenterNet on a NVIDIA Tesla P100 GPU. We obtain that the average inference time of CornerNet511-104 (means that the resolution of input images is 511X511 and the backbone is Hourglass-104) is 300ms per image and that of CenterNet511-104 is 340ms. Meanwhile, using the Hourglass-52 backbone can speed up the inference speed. Our CenterNet511-52 takes an average of 270ms to process per image, which is faster and more accurate than CornerNet511-104.

Preparetion

Please first install Anaconda and create an Anaconda environment using the provided package list.

conda create --name CenterNet --file conda_packagelist.txt

After you create the environment, activate it.

source activate CenterNet

Compiling Corner Pooling Layers

cd <CenterNet dir>/models/py_utils/_cpools/
python setup.py install --user

Compiling NMS

cd <CenterNet dir>/external
make

Installing MS COCO APIs

cd <CenterNet dir>/data/coco/PythonAPI
make

Downloading MS COCO Data

Download the training/validation split we use in our paper from here (originally from Faster R-CNN)
Unzip the file and place annotations under <CenterNet dir>/data/coco
Download the images (2014 Train, 2014 Val, 2017 Test) from here
Create 3 directories, trainval2014, minival2014 and testdev2017, under <CenterNet dir>/data/coco/images/
Copy the training/validation/testing images to the corresponding directories according to the annotation files

Training and Evaluation

To train CenterNet-104:

python train.py CenterNet-104

We provide the configuration file (CenterNet-104.json) and the model file (CenterNet-104.py) for CenterNet in this repo.

We also provide a trained model for CenterNet-104, which is trained for 480k iterations using 8 Tesla V100 (32GB) GPUs. You can download it from [here](https://pan.baidu.com/s/17RvbWaxrvW1kXRuk7XmIfw code:2clj) and put it under <CenterNet dir>/cache/nnet (You may need to create this directory by yourself if it does not exist). If you want to train you own CenterNet, please adjust the batch size in CenterNet-104.json to accommodate the number of GPUs that are available to you.

To use the trained model:

python test.py CenterNet-104 --testiter 480000 --split <split>

To train CenterNet-52:

python train.py CenterNet-52

We provide the configuration file (CenterNet-52.json) and the model file (CenterNet-52.py) for CenterNet in this repo.

We also provide a trained model for CenterNet-52, which is trained for 480k iterations using 8 Tesla V100 (32GB) GPUs. You can download it from [here](https://pan.baidu.com/s/1Ltig0csUPp4T5HA4BjHikA code:ed0y) and put it under <CenterNet dir>/cache/nnet (You may need to create this directory by yourself if it does not exist). If you want to train you own CenterNet, please adjust the batch size in CenterNet-52.json to accommodate the number of GPUs that are available to you.