GluonCV: a Deep Learning Toolkit for Computer Vision

GluonCV provides implementations of state-of-the-art (SOTA) deep learning algorithms in computer vision. It aims to help engineers, researchers, and students quickly prototype products, validate new ideas and learn computer vision.

GluonCV features:

1. training scripts that reproduce SOTA results reported in latest papers,

2. a large set of pre-trained models,

3. carefully designed APIs and easy to understand implementations,

4. community support.

Demo

Check the HD video at Youtube or Bilibili.

Supported Applications

Application	Illustration	Available Models
Image Classification: recognize an object in an image.		50+ models, including ResNet, MobileNet, DenseNet, VGG, ...
Object Detection: detect multiple objects with their bounding boxes in an image.		Faster RCNN, SSD, Yolo-v3
Semantic Segmentation: associate each pixel of an image with a categorical label.		FCN, PSP, DeepLab v3
Instance Segmentation: associate each pixel of an image with an instance label.		Mask RCNN
Pose Estimation: detect human pose from images.		Simple Pose
Video Action Recognition: recognize human actions in a video.		TSN, I3D, Non-local, SlowFast

Installation

Select your preferences and run the install command.

OS: Linux macOS Windows

Version: Stable Nightly Source

Stable Release.

Nightly build with latest features.

Install GluonCV from source.

Backend: Native CUDA MKL-DNN CUDA + MKL-DNN

Build-in backend for CPU.

Required to run on Nvidia GPUs.

Accelerate Intel CPU performance.

Enable both Nvidia GPUs and Intel CPU acceleration.

Prerequisites:

• Requires pip >= 9..

• Note that you can install the extra optional requirements all together by replacing “pip install gluoncv” with “pip install gluoncv[full]”.

• Nightly build provides latest features for enthusiasts.

Command:

# for mxnet
pip install --upgrade mxnet
# for pytorch
pip install torch==1.6.0+cpu torchvision==0.7.0+cpu -f https://download.pytorch.org/whl/torch_stable.html

pip install --upgrade gluoncv

# Here we assume CUDA 10.2 is installed. You can change the number
# according to your own CUDA version.

# for mxnet
pip install --upgrade mxnet-cu102
# for pytorch
pip install torch==1.6.0 torchvision==0.7.0

pip install --upgrade gluoncv

pip install --upgrade mxnet-mkl gluoncv

# Here we assume CUDA 10.2 is installed. You can change the number
# according to your own CUDA version.
pip install --upgrade mxnet-cu102mkl gluoncv

# for mxnet
pip install --upgrade mxnet -f https://dist.mxnet.io/python/all
# for pytorch
pip install --pre torch torchvision torchaudio -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html

pip install --pre --upgrade gluoncv

# mxnet
pip install --upgrade mxnet-cu102 -f https://dist.mxnet.io/python/all
# pytorch
pip install --pre torch torchvision torchaudio -f https://download.pytorch.org/whl/nightly/cu102/torch_nightly.html

pip install --pre --upgrade gluoncv

pip install --pre --upgrade gluoncv
pip install --upgrade mxnet-mkl -f https://dist.mxnet.io/python/all

pip install --pre --upgrade gluoncv
pip install --upgrade mxnet-cu102mkl -f https://dist.mxnet.io/python/all

# mxnet
pip install --upgrade mxnet -f https://dist.mxnet.io/python/all
# pytorch
pip install --pre torch torchvision torchaudio -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html

git clone https://github.com/dmlc/gluon-cv
cd gluon-cv && python setup.py install --user

# mxnet
pip install --upgrade mxnet-cu100 -f https://dist.mxnet.io/python/all
# pytorch
pip install --pre torch torchvision torchaudio -f https://download.pytorch.org/whl/nightly/cu102/torch_nightly.html

git clone https://github.com/dmlc/gluon-cv
cd gluon-cv && python setup.py install --user

pip install --upgrade mxnet-mkl -f https://dist.mxnet.io/python/all
git clone https://github.com/dmlc/gluon-cv
cd gluon-cv && python setup.py install --user

pip install --upgrade mxnet-cu102mkl -f https://dist.mxnet.io/python/all
git clone https://github.com/dmlc/gluon-cv
cd gluon-cv && python setup.py install --user

Check Installation for more installation instructions and options.

A Quick Example

Object Detection Demo

New to Deep Learning or CV?

For background knowledge of deep learning or CV, please refer to the open source book Dive into Deep Learning.

Adoptions

Companies and organizations using GluonCV: