README_en.md

简体中文 | English

PaddleClas

Introduction

PaddleClas is an image classification and image recognition toolset for industry and academia, helping users train better computer vision models and apply them in real scenarios.

PP-ShiTuV2	PULC: Practical Ultra Light-weight image Classification solutions

📣 Recent updates

• 🔥️ Release PP-ShiTuV2, recall1 is improved by nearly 8 points, covering 20+ recognition scenarios, with index management tool and Android Demo for better experience.

• 2022.6.15 Release Practical Ultra Light-weight image Classification solutions. PULC models inference within 3ms on CPU devices, with accuracy on par with SwinTransformer. We also release 9 practical classification models covering pedestrian, vehicle and OCR scenario.

• 2022.4.21 Added the related code of the CVPR2022 oral paper MixFormer.

• 2021.09.17 Add PP-LCNet series model developed by PaddleClas, these models show strong competitiveness on Intel CPUs. For the introduction of PP-LCNet, please refer to paper or PP-LCNet model introduction. The metrics and pretrained model are available here.

• 2021.06.29 Add Swin-transformer) series model，Highest top1 acc on ImageNet1k dataset reaches 87.2%, training, evaluation and inference are all supported. Pretrained models can be downloaded here.

• 2021.06.16 PaddleClas release/2.2. Add metric learning and vector search modules. Add product recognition, animation character recognition, vehicle recognition and logo recognition. Added 30 pretrained models of LeViT, Twins, TNT, DLA, HarDNet, and RedNet, and the accuracy is roughly the same as that of the paper.

• more

🌟 Features

PaddleClas release PP-HGNet、PP-LCNetv2、 PP-LCNet and Simple Semi-supervised Label Distillation algorithms, and support plenty of image classification and image recognition algorithms.Based on th algorithms above, PaddleClas release PP-ShiTu image recognition system and Practical Ultra Light-weight image Classification solutions.

Welcome to Join the Technical Exchange Group

• You can also scan the QR code below to join the PaddleClas QQ group and WeChat group (add and replay "C") to get more efficient answers to your questions and to communicate with developers from all walks of life. We look forward to hearing from you.

Quick Start

Quick experience of PP-ShiTu image recognition system：Link

PP-ShiTuV2 Android Demo

Quick experience of Practical Ultra Light-weight image Classification models：Link

Tutorials

• Install Paddle
• Install PaddleClas Environment
• PP-ShiTuV2 Image Recognition Systems Introduction
  • Image Recognition Quick Start
  • 20+ application scenarios
  • Submodule Introduction and Model Training
    • Mainbody Detection
    • Feature Extraction
    • Vector Search
    • Hash Encoding
  • PipeLine Inference and Deployment
    • Python Inference
    • C++ Inference
    • Serving Deployment
    • Lite Deployment
    • Shitu Gallery Manager Tool
• Practical Ultra Light-weight image Classification solutions
  • PULC Quick Start
  • PULC Model Zoo
    • PULC Classification Model of Someone or Nobody
    • PULC Recognition Model of Person Attribute
    • PULC Classification Model of Wearing or Unwearing Safety Helmet
    • PULC Classification Model of Traffic Sign
    • PULC Recognition Model of Vehicle Attribute
    • PULC Classification Model of Containing or Uncontaining Car
    • PULC Classification Model of Text Image Orientation
    • PULC Classification Model of Textline Orientation
    • PULC Classification Model of Language
• PP Series Backbone
  • PP-HGNet
  • PP-LCNet
  • PP-LCNetv2
• Introduction to Image Recognition Systems
• Image Recognition Demo images
• PULC demo images
• Algorithms Introduction
  • Backbone Network and Pre-trained Model Library
  • Mainbody Detection
  • Feature Learning
  • Vector Search
• Inference Model Prediction
  • Python Inference
  • C++ Classfication Inference
• Model Deploy (only support classification for now, recognition coming soon)
  • Hub Serving Deployment
  • Mobile Deployment
  • Inference Using whl
• Advanced Tutorial
  • Knowledge Distillation
  • Model Quantization
  • Data Augmentation
• License
• Contribution

Introduction to Practical Ultra Light-weight image Classification solutions

PULC solutions consists of PP-LCNet light-weight backbone, SSLD pretrained models, Ensemble of Data Augmentation strategy and SKL-UGI knowledge distillation. PULC models inference within 3ms on CPU devices, with accuracy comparable with SwinTransformer. We also release 9 practical models covering pedestrian, vehicle and OCR.

Introduction to Image Recognition Systems

PP-ShiTuV2 is a practical lightweight general image recognition system, which is mainly composed of three modules: mainbody detection model, feature extraction model and vector search tool. The system adopts a variety of strategies including backbone network, loss function, data augmentations, optimal hyperparameters, pre-training model, model pruning and quantization. Compared to V1, PP-ShiTuV2, Recall1 is improved by nearly 8 points. For more details, please refer to PP-ShiTuV2 introduction. For a new unknown category, there is no need to retrain the model, just prepare images of new category, extract features and update retrieval database and the category can be recognised.

PP-ShiTuV2 Demo images

• Drinks recognition

• Product recognition

• Cartoon character recognition

• Logo recognition

• Car recognition

PULC demo images

License

PaddleClas is released under the Apache 2.0 license Apache 2.0 license

Contribution

Contributions are highly welcomed and we would really appreciate your feedback!!

• Thank nblib to fix bug of RandErasing.
• Thank chenpy228 to fix some typos PaddleClas.
• Thank jm12138 to add ViT, DeiT models and RepVGG models into PaddleClas.
• Thank FutureSI to parse and summarize the PaddleClas code.