Few-Shot Object Detection - Hands-on MMFewShot

OpenMMLab是香港中文大學多媒體實驗室(MMLab)與商湯科技合作建立的AI開源算法平台。在2021年OpenMMLab為發展Few-Shot Classification 與Few-Shot Object Detection這兩個題目開啟了MMFewShot這個新專案。詳見官網。

Contents of MMFewShot

• Few-shot classification/object detection algorithms
• Inheritable modules and configure file system
• Pre-trained model weights

Tutorials

🔥Introduction and Meta-Inference

🔥Meta-Training

Official Documents

Official Introduction to mmfewshot

• Installation
• Concepts of FSL.
• Basic usage of mmfewshot

Model Zoo

Algorithms that MMFewShot implemented, with downloadable pre-trained weight.

Classification

• Baseline (ICLR'2019)
• Baseline++ (ICLR'2019)
• NegMargin (ECCV'2020)
• MatchingNet (NeurIPS'2016)
• ProtoNet (NeurIPS'2017)
• RelationNet (CVPR'2018)
• MetaBaseline (ICCV'2021)
• MAML (ICML'2017)

Detection

• TFA (ICML'2020)
• FSCE (CVPR'2021)
• AttentionRPN (CVPR'2020)
• MetaRCNN (ICCV'2019)
• FSDetView (ECCV'2020)
• MPSR (ECCV'2020)

Citation

@misc{mmfewshot2021,
    title={OpenMMLab Few Shot Learning Toolbox and Benchmark},
    author={mmfewshot Contributors},
    howpublished = {\url{https://github.com/open-mmlab/mmfewshot}},
    year={2021}
}

@article{wang2020few,
    title={Frustratingly Simple Few-Shot Object Detection},
    author={Wang, Xin and Huang, Thomas E. and  Darrell, Trevor and Gonzalez, Joseph E and Yu, Fisher}
    booktitle = {International Conference on Machine Learning (ICML)},
    month = {July},
    year = {2020}
}

Acknowledgements

• https://github.com/open-mmlab/mmfewshot
  • Official repository of mmfewshot
• https://github.com/ucbdrive/few-shot-object-detection
  • Method of data preparation
• https://app.roboflow.com/
  • Data source, method of data preparation