https://github.com/SsisyphusTao/Object-Detection-Knowledge-Distillation

An Object Detection Knowledge Distillation framework powered by pytorch, now having SSD and yolov5.
https://github.com/SsisyphusTao/Object-Detection-Knowledge-Distillation

deep-learning knowledge-distillation object-detection pytorch ssd yolov5

Last synced: about 1 month ago
JSON representation

An Object Detection Knowledge Distillation framework powered by pytorch, now having SSD and yolov5.

• Host: GitHub
• URL: https://github.com/SsisyphusTao/Object-Detection-Knowledge-Distillation
• Owner: SsisyphusTao
• License: gpl-3.0
• Created: 2020-05-10T09:09:53.000Z (about 4 years ago)
• Default Branch: dev
• Last Pushed: 2022-02-07T12:06:47.000Z (over 2 years ago)
• Last Synced: 2024-05-12T22:21:09.331Z (about 2 months ago)
• Topics: deep-learning, knowledge-distillation, object-detection, pytorch, ssd, yolov5
• Language: Python
• Homepage:
• Size: 628 KB
• Stars: 212
• Watchers: 6
• Forks: 51
• Open Issues: 12
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Lists

• awesome-knowledge-distillation - Knowledge Distillation on SSD
• awesome-yolo-object-detection - SsisyphusTao/Object-Detection-Knowledge-Distillation - Detection-Knowledge-Distillation?style=social"/> : An Object Detection Knowledge Distillation framework powered by pytorch, now having SSD and yolov5. (Lighter and Deployment Frameworks)

README

        
# Object Detection Knowledge Distillation(ODKD)

![version](https://img.shields.io/badge/version-beta_0.0.1-brightgreen.svg)

![coverage](https://codecov.io/github/SsisyphusTao/SSD-Knowledge-Distillation/coverage.svg?branch=dev)

The function of this branch is not complete. For [ssd](https://github.com/SsisyphusTao/SSD-Knowledge-Distillation/tree/mbv2-lite) and [yolov5](https://github.com/SsisyphusTao/Object-Detection-Knowledge-Distillation/tree/yolov5) distillation, checking other branches. 

Release edition is coming Soon...

## Update

1. The first edition is the refactor of branch [mbv2-lite](https://github.com/SsisyphusTao/SSD-Knowledge-Distillation/tree/mbv2-lite), which is an implementation of Chen, G. et al. (2017) [‘Learning efficient object detection models with knowledge distillation’](http://papers.nips.cc/paper/6676-learning-efficient-object-detection-models-with-knowledge-distillation.pdf) with **SSD-lite** structure.

2. Replace part of code with pytorch api which has same functionality.

3. Very friendly beginner guidance.

4. System Architecture

![odkd](http://assets.processon.com/chart_image/6198ae621efad406f87a16ec.png)

## Useage

```

$ python setup.py install --user

$ odkd-train ./training_config.yml -t

$ odkd-train training_config.yml

or

$ python -m torch.distributed.launch --nproc_per_node=2 `which odkd-train` training_config.yml

$ odkd-eval ${CHECKPOINTS_PATH}/${RUN_INDEX}/config.yml

```

## TODO

- Evaluation Module

- LOG Module

- Coco dataset support

- Yolov5 distillation