According to the "Sound of the Chinese Academy of Sciences", the Center for Intelligent Perception and Computation of the Institute of Automation of the Chinese Academy of Sciences recently proposed the High Fidelity Pose Invariant Model (HF-PIM) based on the generation of the confrontation network. Overcome the most classic pose inconsistency in face recognition tasks. The experimental results show that the visual and quantitative performance indicators of the proposed method on the benchmark dataset are better than the best methods based on the anti-generation network. In addition, the resolution of the generated image supported by HF-PIM has doubled the original method. The paper was included in the Conference on Neural Information Processing Systems (NIPS). In order to solve some limitations in the previous work, the author of the paper introduced a correlated dense dense correlation field that can reflect the point-to-point between the 3D face model and the 2D face image, so that the network can be in the 2D image. Under the guidance of the three-dimensional face information; and designed a new texture warping process, can effectively map the face texture to the image domain, while maintaining the semantic information of the input to the maximum extent. And proposed a learning process against the residual dictionary, so that face texture features can be learned more effectively without relying on three-dimensional data. The experimental results show that the visual and quantitative performance indicators of the proposed method on the benchmark dataset are better than the best methods based on the anti-generation network. In addition, the resolution of the generated image supported by HF-PIM has doubled the original method. The paper was included in the Conference on Neural Information Processing Systems (NIPS). It is understood that the creation of the confrontation network is a revolutionary new development after the deep neural network. It has been rated as the "Top Ten Global Breakthrough Technologies" by the MIT Technology Review in 2018, and competed through two AI systems. Confrontation, maximizing the process of accelerating machine learning, and thus giving the imagination of machine intelligence never before. Cnc Steel Parts,Cnc Machine Accessories,Cnc Steel Turning Parts,Cnc Machining Steel Parts Huaying Chuang Precision Machinery Co., Ltd. , https://www.hycmetal.com