外表互换:联合判别和生成学习用于REID:Joint Discriminative and Generative Learning for Person Re-identification

https://zhuanlan.zhihu.com/p/66408913

market1501:mAp86,rank186
外表互换:联合判别和生成学习用于REID:Joint Discriminative and Generative Learning for Person Re-identification_第1张图片

1 出处2019 NVIDIA/悉尼科技大学 郑哲东

2 key,图像分成两个空间space,一个appearance为颜色、纹理等;另一个是structure空间,body parts

3 key:N副图做NN配对交换appearance和structure生成NN张图再进行end2end学习

外表互换:联合判别和生成学习用于REID:Joint Discriminative and Generative Learning for Person Re-identification_第2张图片

你可能感兴趣的:(#,可以)