kyxt
/
code_test
Not watched
1
0
Fork 0
Code
Releases 0 Wiki evaluate Activity
Issues 0 Pull Requests 0 Datasets Model Cloudbrain HPC
You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
6 Commits
1 Branch
316 MB
18 Download
Tree: eda60a3b22
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from 'eda60a3b22'
${ noResults }
code_test/configs/yolo/README.md

README.md 4.3 kB
Raw Normal View History

first commit
2 years ago
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
 123456789101112131415161718192021222324252627282930313233343536373839404142434445 
  1. # YOLOv3

  2. 

  3. ## Introduction

  4. 

  5. <!-- [ALGORITHM] -->

  6. 

  7. ```latex

  8. @misc{redmon2018yolov3,

  9.     title={YOLOv3: An Incremental Improvement},

  10.     author={Joseph Redmon and Ali Farhadi},

  11.     year={2018},

  12.     eprint={1804.02767},

  13.     archivePrefix={arXiv},

  14.     primaryClass={cs.CV}

  15. }

  16. ```

  17. 

  18. ## Results and Models

  19. 

  20. |    Backbone     |  Scale  | Lr schd | Mem (GB) | Inf time (fps) | box AP | Config | Download  |

  21. | :-------------: | :-----: | :-----: | :------: | :------------: | :----: | :------: | :--------: |

  22. |   DarkNet-53    |   320   |   273e  |   2.7    |      63.9      |  27.9  | [config](https://github.com/open-mmlab/mmdetection/tree/master/configs/yolo/yolov3_d53_320_273e_coco.py) | [model](https://download.openmmlab.com/mmdetection/v2.0/yolo/yolov3_d53_320_273e_coco/yolov3_d53_320_273e_coco-421362b6.pth) &#124; [log](https://download.openmmlab.com/mmdetection/v2.0/yolo/yolov3_d53_320_273e_coco/yolov3_d53_320_273e_coco-20200819_172101.log.json) |

  23. |   DarkNet-53    |   416   |   273e  |   3.8    |      61.2      |  30.9  | [config](https://github.com/open-mmlab/mmdetection/tree/master/configs/yolo/yolov3_d53_mstrain-416_273e_coco.py) | [model](https://download.openmmlab.com/mmdetection/v2.0/yolo/yolov3_d53_mstrain-416_273e_coco/yolov3_d53_mstrain-416_273e_coco-2b60fcd9.pth) &#124; [log](https://download.openmmlab.com/mmdetection/v2.0/yolo/yolov3_d53_mstrain-416_273e_coco/yolov3_d53_mstrain-416_273e_coco-20200819_173424.log.json) |

  24. |   DarkNet-53    |   608   |   273e  |   7.4    |      48.1      |  33.7  | [config](https://github.com/open-mmlab/mmdetection/tree/master/configs/yolo/yolov3_d53_mstrain-608_273e_coco.py) | [model](https://download.openmmlab.com/mmdetection/v2.0/yolo/yolov3_d53_mstrain-608_273e_coco/yolov3_d53_mstrain-608_273e_coco_20210518_115020-a2c3acb8.pth) &#124; [log](https://download.openmmlab.com/mmdetection/v2.0/yolo/yolov3_d53_mstrain-608_273e_coco/yolov3_d53_mstrain-608_273e_coco_20210518_115020.log.json) |

  25. 

  26. ## Mixed Precision Training

  27. 

  28. We also train YOLOv3 with mixed precision training.

  29. 

  30. |    Backbone     |  Scale  | Lr schd | Mem (GB) | Inf time (fps) | box AP | Config | Download  |

  31. | :-------------: | :-----: | :-----: | :------: | :------------: | :----: | :------: | :--------: |

  32. |   DarkNet-53    |   608   |   273e  |   4.7    |      48.1      |  33.8  | [config](https://github.com/open-mmlab/mmdetection/tree/master/configs/yolo/yolov3_d53_fp16_mstrain-608_273e_coco.py) | [model](https://download.openmmlab.com/mmdetection/v2.0/yolo/yolov3_d53_fp16_mstrain-608_273e_coco/yolov3_d53_fp16_mstrain-608_273e_coco_20210517_213542-4bc34944.pth) &#124; [log](https://download.openmmlab.com/mmdetection/v2.0/yolo/yolov3_d53_fp16_mstrain-608_273e_coco/yolov3_d53_fp16_mstrain-608_273e_coco_20210517_213542.log.json) |

  33. 

  34. ## Lightweight models

  35. 

  36. |    Backbone     |  Scale  | Lr schd | Mem (GB) | Inf time (fps) | box AP | Config   | Download   |

  37. | :-------------: | :-----: | :-----: | :------: | :------------: | :----: | :------: | :--------: |

  38. |   MobileNetV2   |   416   |   300e  |   5.3    |                |  23.9  | [config](https://github.com/open-mmlab/mmdetection/tree/master/configs/yolo/yolov3_mobilenetv2_mstrain-416_300e_coco.py) | [model](https://download.openmmlab.com/mmdetection/v2.0/yolo/yolov3_mobilenetv2_mstrain-416_300e_coco/yolov3_mobilenetv2_mstrain-416_300e_coco_20210718_010823-f68a07b3.pth) &#124; [log](https://download.openmmlab.com/mmdetection/v2.0/yolo/yolov3_mobilenetv2_mstrain-416_300e_coco/yolov3_mobilenetv2_mstrain-416_300e_coco_20210718_010823.log.json) |

  39. |   MobileNetV2   |   320   |   300e  |   3.2    |                |  22.2  | [config](https://github.com/open-mmlab/mmdetection/tree/master/configs/yolo/yolov3_mobilenetv2_320_300e_coco.py) | [model](https://download.openmmlab.com/mmdetection/v2.0/yolo/yolov3_mobilenetv2_320_300e_coco/yolov3_mobilenetv2_320_300e_coco_20210719_215349-d18dff72.pth) &#124; [log](https://download.openmmlab.com/mmdetection/v2.0/yolo/yolov3_mobilenetv2_320_300e_coco/yolov3_mobilenetv2_320_300e_coco_20210719_215349.log.json) |

  40. 

  41. Notice: We reduce the number of channels to 96 in both head and neck. It can reduce the flops and parameters, which makes these models more suitable for edge devices.

  42. 

  43. ## Credit

  44. 

  45. This implementation originates from the project of Haoyu Wu(@wuhy08) at Western Digital.

No Description

Contributors (2)
All