Welcome to CVNets documentation!

CVNets is an open-source library for training deep neural networks for visual recognition tasks, including classification, detection, and segmentation.

CVNets supports image and video understanding tools, including data loading, data transformations, novel data sampling methods, and implementations of several state-of-the-art networks.

Our source code is available on Github .

Table of Contents

• Getting Started
  • Installation
  • Config Files: Introduction and Walkthrough
  • Directory Structure
  • Training Classification Models on the ImageNet dataset
  • Object detection using SSDLite on MS-COCO
  • Semantic Segmentation using DeepLabv3
• Sample Recipes
  • Training ResNets on the ImageNet dataset
  • Training MobileNets on the ImageNet dataset
  • Training MobileViT on the ImageNet dataset
  • Training MobileViTv2 Models
• How To
  • Converting models trained using CVNets to CoreML
  • How to Create a New Dataset Type
• Data Samplers
  • Single-scale with fixed batch size (SSc-FBS)
  • Multi-scale with fixed batch size (MSc-FBS)
  • Multi-scale with variable batch size (MSc-VBS):
  • Data Sampler Objects
• Model Zoo
  • Classification (ImageNet-1k)
  • Object Detection (MS-COCO)
  • Segmentation
  • Video Classification (Kinetics-400)

Citation

If you find CVNets useful, please cite the following papers:

@inproceedings{mehta2022mobilevit,
    title={MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer},
    author={Sachin Mehta and Mohammad Rastegari},
    booktitle={International Conference on Learning Representations},
    year={2022}
}

@inproceedings{mehta2022cvnets,
    author = {Mehta, Sachin and Abdolhosseini, Farzad and Rastegari, Mohammad},
    title = {CVNets: High Performance Library for Computer Vision},
    year = {2022},
    booktitle = {Proceedings of the 30th ACM International Conference on Multimedia},
    series = {MM '22}
}

Indices and tables

• Index

• Module Index