With the proposed PAGAN, the face recognition accuracy with synthesized images has increased 0.21% and the image quality rating has increased around 5%, which proves the effectiveness and validity of proposed method. Extensive experiments demonstrate that the proposed method outperforms state-of-the-art methods not only in the accuracy of age classification but also in the image quality. The proposed face aging framework with PAGAN is a combination of age estimation, identity preservation, and image de-noising. Face images are featured by age, identity, and fine-grained pixel-value to ensure the quality, which is a typical multi-task learning problem. To meet this challenge, we propose a face aging framework named as Pixel-level Alignment GAN, PAGAN, to synthesize faces of different age groups. However, there is still a huge gap between the synthesized face image and the real face in terms of quality and consistency due to identity ambiguity and image distortion caused by existing face aging methods.
Face morph age progression online verification#
įace aging is of great significance in cross-time identity verification problem. We further validate MTLFace on two popular general face recognition datasets, obtaining competitive performance on face recognition in the wild. Extensive experimental results on five benchmark cross-age datasets demonstrate that MTLFace yields superior performance for both AIFR and FAS. Furthermore, to advance both AIFR and FAS, we collect and release a large cross-age face dataset with age and gender annotations, and a new benchmark specifically designed for tracing long-missing children. Benefiting from the proposed multi-task framework, we then leverage those high-quality synthesized faces from FAS to further boost AIFR via a novel selective fine-tuning strategy. Unlike the conventional one-hot encoding that achieves group-level FAS, we propose a novel identity conditional module to achieve identity-level FAS, which can improve the age smoothness of synthesized faces through a weight-sharing strategy. Specifically, we propose an attention-based feature decomposition to decompose the mixed face features into two uncorrelated components - identity- and age-related features - in a spatially constrained way. Therefore, we propose a unified, multi-task framework to jointly handle these two tasks, termed MTLFace, which can learn the age-invariant identity-related representation for face recognition while achieving pleasing face synthesis for model interpretation. However, AIFR lacks visual results for model interpretation and FAS compromises downstream recognition due to artifacts. To minimize the impact of age variation on face recognition, age-invariant face recognition (AIFR) extracts identity-related discriminative features by minimizing the correlation between identity- and age-related features while face age synthesis (FAS) eliminates age variation by converting the faces in different age groups to the same group.
Face morph age progression online code#
The source code and datasets are available at Extensive experimental results on five benchmark cross-age datasets demonstrate that MTLFace yields superior performance than state-of-the-art methods for both AIFR and FAS. Specifically, we propose an attention-based feature decomposition to decompose the mixed face features into two uncorrelated components-identity- and age-related features-in a spatially constrained way.
0 kommentar(er)
