• MindCV
• • Introduction
  • • Benchmark Results
  • Installation
  • • Dependency
    • Install with PyPI
    • Install from Source
    • • Notes:
  • Get Started
  • • Hands-on Tutorial
    • Training and Validation
  • Tutorials
  • Model List
  • Supported Algorithms
  • Notes
  • • What is New
    • How to Contribute
    • License
    • Acknowledgement
    • Citation

MindCV

Build Type	Linux	MacOS	Windows
Build Status			Not tested

English | 中文

Introduction |
Installation |
Get Started |
Tutorials |
Model List |
Supported Algorithms |
Notes

Introduction

MindCV is an open source toolbox for computer vision research and development based on MindSpore. It collects a series of classic and SoTA vision models, such as ResNet and SwinTransformer, along with their pretrained weights. SoTA methods such as AutoAugment are also provided for performance improvement. With the decoupled module design, it is easy to apply or adapt MindCV to your own CV tasks.

Major Features

• Easy-to-Use. MindCV decomposes the vision framework into various configurable components. It is easy to customize your data pipeline, models, and learning pipeline with MindCV:

>>> import mindcv 
# create a dataset
>>> dataset = mindcv.create_dataset('cifar10', download=True)
# create a model
>>> network = mindcv.create_model('resnet50', pretrained=True)

Users can customize and launch their transfer learning or training task in one command line.

# transfer learning in one command line
>>> !python train.py --model=swin_tiny --pretrained --opt=adamw --lr=0.001 --data_dir={data_dir} 

• State-of-The-Art. MindCV provides various CNN-based and Transformer-based vision models including SwinTransformer. Their pretrained weights and performance reports are provided to help users select and reuse the right model:
  __
• Flexibility and efficiency. MindCV is bulit on MindSpore which is an efficent DL framework that can run on different hardward platforms (GPU/CPU/Ascend). It supports both graph mode for high efficiency and pynative mode for flexibity.

Benchmark Results

The performance of the models trained with MindCV is summarized in benchmark_results.md, where the training recipes and weights are both available.

Model introduction and training details can be viewed in each subfolder under configs.

Installation

Dependency

• mindspore >= 1.8.1
• numpy >= 1.17.0
• pyyaml >= 5.3
• tqdm
• openmpi 4.0.3 (for distributed mode)

To install the dependency, please run

pip install -r requirements.txt

MindSpore can be easily installed by following the official instruction where you can select your hardware platform for the best fit. To run in distributed mode, openmpi is required to install.

The following instructions assume the desired dependency is fulfilled.

Install with PyPI

The released version of MindCV can be installed via PyPI as follows:

pip install mindcv

Install from Source

The latest version of MindCV can be installed as follows:

pip install git+https://github.com/mindspore-lab/mindcv.git

Notes:

• MindCV can be installed on Linux and Mac but not on Windows currently.

Get Started

Hands-on Tutorial

To get started with MindCV, please see the transfer learning tutorial, which will give a quick tour on each key component and the train/validate/predict pipelines in MindCV.

Below is a few code snippets for your taste.

>>> import mindcv 
# List and find a pretrained vision model 
>>> mindcv.list_models("swin*", pretrained=True)
['swin_tiny']
# Create the model object
>>> network = mindcv.create_model('swin_tiny', pretrained=True)
# Validate its accuracy
>>> !python validate.py --model=swin_tiny --pretrained --dataset=imagenet --val_split=validation
{'Top_1_Accuracy': 0.808343989769821, 'Top_5_Accuracy': 0.9527253836317136, 'loss': 0.8474242982580839}

Image classification demo

Infer the input image with a pretrained SoTA model,

>>> !python infer.py --model=swin_tiny --image_path='./tutorials/data/test/dog/dog.jpg'
{'Labrador retriever': 0.5700152, 'golden retriever': 0.034551315, 'kelpie': 0.010108651, 'Chesapeake Bay retriever': 0.008229004, 'Walker hound, Walker foxhound': 0.007791956}

The top-1 prediction result is labrador retriever (拉布拉多犬), which is the breed of this cut dog.

Training and Validation

It is easy to train your model on standard datasets or your own dataset with MindCV. Model training, transfer learning, or evaluaiton can be done using one or a few line of code with flexible configuration.

• Standalone Training

It is easy to do model training with train.py. Here is an example for training a DenseNet on CIFAR10 dataset using one computing device (i.e., standalone GPU).

python train.py --model=resnet50 --dataset=cifar10 --dataset_download

For more parameter description, please run `python train.py --help'. You can define change model, optimizer, and other hyper-parameters easily.

Validation while training. To track the validation accuracy change during traing, please enable --val_while_train, for example

python train.py --model=resnet50 --dataset=cifar10 \
		--val_while_train --val_split=test --val_interval=1

The training loss and validation accuracy for each epoch will be saved in {ckpt_save_dir}/results.log.

Resume training. To resume training, please specify --ckpt_path for the checkpoint where you want to resume and --ckpt_save_dir. The optimizer state including learning rate of the last epoch will also be recovered.

python train.py --model=resnet50 --dataset=cifar10 \
		--ckpt_save_dir=checkpoints --ckpt_path=checkpoints/resnet50_30-100.ckpt

• Distributed Training

For large datasets like ImageNet, it is necessary to do training in distributed mode on multiple devices, which is well supported in MindCV. The following script is an example for training DenseNet121 on ImageNet with 4 GPUs.

export CUDA_VISIBLE_DEVICES=0,1,2,3  # suppose there are 4 GPUs
mpirun --allow-run-as-root -n 4 python train.py --distribute \
	--model=densenet121 --dataset=imagenet --data_dir=./datasets/imagenet   

• Configuration with Yaml

You can configure your model and other components either by specifying external parameters or by using a yaml config file. Here is an example for training using a preset yaml file.

mpirun --allow-run-as-root -n 4 python train.py -c configs/squeezenet/squeezenet_1.0_gpu.yaml    

Well-defined config files for training SoTA models can be seen in the configs folder, along with their performance reported on ImageNet dataset.

• Validation

It is easy to validate a trained model with validate.py.

# validate a trained checkpoint
python validate.py --model=resnet50 --dataset=imagenet --val_split=validation --ckpt_path='./ckpt/densenet121-best.ckpt' 

• Pynative mode with ms_function (Experimental)

By default, the training pipeline (train.py) is run in graph mode, which is optimized for efficienty and speed but may not be flexible enough for debugging. You may alter the parameter --mode to switch to pure pynative mode for debugging purpose.

Pynative mode with ms_function is a mixed mode for comprising flexibity and efficiency in MindSpore. To switch to pynative mode with ms_function, please use train_with_func.py instead, for example:

python train_with_func.py --model=resnet50 --dataset=cifar10 --dataset_download  --epoch_size=10  

Note: train_with_func.py is an experimental training pipeline and is less stable than train.py. It will be improved after MindSpore 2.0 released.

For more examples, see examples/scripts.

Tutorials

We provide the following jupyter notebook tutorials to help users learn to use MindCV.

• Learn about configs
• Inference with a pretrained model
• Finetune a pretrained model on custom datasets
• Customize your model //coming soon
• Optimizing performance for vision transformer //coming soon
• Deployment demo

Model List

Currently, MindCV supports the model families listed below. More models with pretrained weights are under development and will be released soon.

Supported models

• Big Transfer ResNetV2 (BiT) - https://arxiv.org/abs/1912.11370
• ConvNeXt - https://arxiv.org/abs/2201.03545
• ConViT (Soft Convolutional Inductive Biases Vision Transformers)- https://arxiv.org/abs/2103.10697
• DenseNet - https://arxiv.org/abs/1608.06993
• DPN (Dual-Path Network) - https://arxiv.org/abs/1707.01629
• EfficientNet (MBConvNet Family) https://arxiv.org/abs/1905.11946
• EfficientNet V2 - https://arxiv.org/abs/2104.00298
• GhostNet - https://arxiv.org/abs/1911.11907
• GoogleNet - https://arxiv.org/abs/1409.4842
• Inception-V3 - https://arxiv.org/abs/1512.00567
• Inception-ResNet-V2 and Inception-V4 - https://arxiv.org/abs/1602.07261
• MNASNet - https://arxiv.org/abs/1807.11626
• MobileNet-V1 - https://arxiv.org/abs/1704.04861
• MobileNet-V2 - https://arxiv.org/abs/1801.04381
• MobileNet-V3 (MBConvNet w/ Efficient Head) - https://arxiv.org/abs/1905.02244
• NASNet - https://arxiv.org/abs/1707.07012
• PNasNet - https://arxiv.org/abs/1712.00559
• PVT (Pyramid Vision Transformer) - https://arxiv.org/abs/2102.12122
• PoolFormer models - https://github.com/sail-sg/poolformer
• RegNet - https://arxiv.org/abs/2003.13678
• RepMLP https://arxiv.org/abs/2105.01883
• RepVGG - https://arxiv.org/abs/2101.03697
• ResNet (v1b/v1.5) - https://arxiv.org/abs/1512.03385
• ResNeXt - https://arxiv.org/abs/1611.05431
• Res2Net - https://arxiv.org/abs/1904.01169
• ReXNet - https://arxiv.org/abs/2007.00992
• ShuffleNet v1 - https://arxiv.org/abs/1707.01083
• ShuffleNet v2 - https://arxiv.org/abs/1807.11164
• SKNet - https://arxiv.org/abs/1903.06586
• SqueezeNet - https://arxiv.org/abs/1602.07360
• Swin Transformer - https://arxiv.org/abs/2103.14030
• VGG - https://arxiv.org/abs/1409.1556
• Visformer - https://arxiv.org/abs/2104.12533
• Vision Transformer (ViT) - https://arxiv.org/abs/2010.11929
• Xception - https://arxiv.org/abs/1610.02357

Please see configs for the details about model performance and pretrained weights.

Supported Algorithms

Supported algorithms

• Augmentation
  • AutoAugment
  • RandAugment
  • Repeated Augmentation
  • RandErasing (Cutout)
  • CutMix
  • Mixup
  • RandomResizeCrop
  • Color Jitter, Flip, etc
• Optimizer
  • Adam
  • Adamw
  • Adan (experimental)
  • AdaGrad
  • LAMB
  • Momentum
  • RMSProp
  • SGD
  • NAdam
• LR Scheduler
  • Warmup Cosine Decay
  • Step LR
  • Polynomial Decay
  • Exponential Decay
• Regularization
  • Weight Decay
  • Label Smoothing
  • Stochastic Depth (depends on networks)
  • Dropout (depends on networks)
• Loss
  • Cross Entropy (w/ class weight and auxilary logit support)
  • Binary Cross Entropy (w/ class weight and auxilary logit support)
  • Soft Cross Entropy Loss (automatically enabled if mixup or label smoothing is used)
  • Soft Binary Cross Entropy Loss (automatically enabled if mixup or label smoothing is used)
• Ensemble
  • Warmup EMA (Exponential Moving Average)

Notes

What is New

• 2022/12/09

1. Support lr warmup for all lr scheduling algorithms besides cosine decay.
2. Add repeated augmentation, which can be enabled by setting --aug_repeats to be a value larger than 1 (typically, 3 or 4 is a common choice).
3. Add EMA.
4. Improve BCE loss to support mixup/cutmix.

• 2022/11/21

1. Add visualization for loss and acc curves
2. Support epochwise lr warmup cosine decay (previous is stepwise)

• 2022/11/09

1. Add 7 pretrained ViT models.
2. Add RandAugment augmentation.
3. Fix CutMix efficiency issue and CutMix and Mixup can be used together.
4. Fix lr plot and scheduling bug.

• 2022/10/12

1. Both BCE and CE loss now support class-weight config, label smoothing, and auxilary logit input (for networks like inception).

• 2022/09/13

1. Add Adan optimizer (experimental)

How to Contribute

We appreciate all kind of contributions including issues and PRs to make MindCV better.

Please refer to CONTRIBUTING.md for the contributing guideline. Please follow the Model Template and Guideline for contributing a model that fits the overall interface :)

License

This project follows the Apache License 2.0 open-source license.

Acknowledgement

MindCV is an open-source project jointly developed by the MindSpore team, Xidian University, and Xi'an Jiaotong University.
Sincere thanks to all participating researchers and developers for their hard work on this project.
We also acknowledge the computing resources provided by OpenI.

Citation

If you find this project useful in your research, please consider citing:

@misc{MindSpore Computer Vision 2022,
    title={{MindSpore Computer  Vision}:MindSpore Computer Vision Toolbox and Benchmark},
    author={MindSpore Vision Contributors},
    howpublished = {\url{https://github.com/mindspore-lab/mindcv/}},
    year={2022}
}