# Distillation(蒸馏) ## YOLOv3模型蒸馏 以YOLOv3-MobileNetV1为例,使用YOLOv3-ResNet34作为蒸馏训练的teacher网络, 对YOLOv3-MobileNetV1结构的student网络进行蒸馏。 COCO数据集作为目标检测任务的训练目标难度更大,意味着teacher网络会预测出更多的背景bbox,如果直接用teacher的预测输出作为student学习的`soft label`会有严重的类别不均衡问题。解决这个问题需要引入新的方法,详细背景请参考论文:[Object detection at 200 Frames Per Second](https://arxiv.org/abs/1805.06361)。 为了确定蒸馏的对象,我们首先需要找到student和teacher网络得到的`x,y,w,h,cls,objness`等Tensor,用teacher得到的结果指导student训练。具体实现可参考[代码](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/ppdet/slim/distill.py) ## FGD模型蒸馏 FGD全称为[Focal and Global Knowledge Distillation for Detectors](https://arxiv.org/abs/2111.11837v1),是目标检测任务的一种蒸馏方法,FGD蒸馏分为两个部分`Focal`和`Global`。`Focal`蒸馏分离图像的前景和背景,让学生模型分别关注教师模型的前景和背景部分特征的关键像素;`Global`蒸馏部分重建不同像素之间的关系并将其从教师转移到学生,以补偿`Focal`蒸馏中丢失的全局信息。试验结果表明,FGD蒸馏算法在基于anchor和anchor free的方法上能有效提升模型精度。 在PaddleDetection中,我们实现了FGD算法,并基于retinaNet算法进行验证,实验结果如下: | algorithm | model | AP | download| |:-:| :-: | :-: | :-:| |retinaNet_r101_fpn_2x | teacher | 40.6 | [download](https://paddledet.bj.bcebos.com/models/retinanet_r101_fpn_2x_coco.pdparams) | |retinaNet_r50_fpn_1x| student | 37.5 |[download](https://paddledet.bj.bcebos.com/models/retinanet_r50_fpn_1x_coco.pdparams) | |retinaNet_r50_fpn_2x + FGD| student | 40.8 |[download](https://paddledet.bj.bcebos.com/models/retinanet_r101_distill_r50_2x_coco.pdparams) | ## LD模型蒸馏 LD全称为[Localization Distillation for Dense Object Detection](https://arxiv.org/abs/2102.12252),将回归框表示为概率分布,把分类任务的KD用在定位任务上,并且使用因地制宜、分而治之的策略,在不同的区域分别学习分类知识与定位知识。在PaddleDetection中,我们实现了LD算法,并基于GFL模型进行验证,实验结果如下: | algorithm | model | AP | download| |:-:| :-: | :-: | :-:| | GFL_ResNet101-vd | teacher | 46.8 | [model](https://paddledet.bj.bcebos.com/models/gfl_r101vd_fpn_mstrain_2x_coco.pdparams), [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/configs/gfl/gfl_r101vd_fpn_mstrain_2x_coco.yml) | | GFL_ResNet18-vd | student | 36.6 | [model](https://paddledet.bj.bcebos.com/models/gfl_r18vd_1x_coco.pdparams), [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/configs/gfl/gfl_r18vd_1x_coco.yml) | | GFL_ResNet18-vd + LD | student | 38.2 | [model](https://bj.bcebos.com/v1/paddledet/models/gfl_slim_ld_r18vd_1x_coco.pdparams), [config1](../../gfl/gfl_slim_ld_r18vd_1x_coco.yml), [config2](./gfl_ld_distill.yml) | ## Citations ``` @article{mehta2018object, title={Object detection at 200 Frames Per Second}, author={Rakesh Mehta and Cemalettin Ozturk}, year={2018}, eprint={1805.06361}, archivePrefix={arXiv}, primaryClass={cs.CV} } @inproceedings{yang2022focal, title={Focal and global knowledge distillation for detectors}, author={Yang, Zhendong and Li, Zhe and Jiang, Xiaohu and Gong, Yuan and Yuan, Zehuan and Zhao, Danpei and Yuan, Chun}, booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition}, pages={4643--4652}, year={2022} } @Inproceedings{zheng2022LD, title={Localization Distillation for Dense Object Detection}, author= {Zheng, Zhaohui and Ye, Rongguang and Wang, Ping and Ren, Dongwei and Zuo, Wangmeng and Hou, Qibin and Cheng, Mingming}, booktitle={CVPR}, year={2022} } ```