我们提出了一种跨摄像头的风格迁移方法，将starGAN应用于摄像头之间

We propose a method of cross-migration style camera, the starGAN style applied to transfer between the camera and soften re-distributed label in recognition training. In this way the entire imaging system to obtain a better model generalization ability.

We propose a style migration method across cameras, apply inggan to the style shift between cameras, and soften the distribution labels in re-recognition training. This method can obtain a better model of generalization ability in the full camera system.

We propose a cross camera style transfer method, which applies stargan to the style transfer between cameras, and softens the distributed tags in the re recognition training. This method can get better generalization ability model in the whole camera system.<br>