利用 Wasserstein 距离的年龄不变面部嵌入（英文版）

英文标题：Age-Invariant Face Embedding using the Wasserstein Distance中文摘要：通过使用多任务学习和 Wasserstein 距离鉴别器来解决面部识别中存在的年龄变化问题，这种方法通过最小化 Jensen-Shannon 差异来最小化年龄和身份嵌入之间的互信息，并在多个年龄变化数据集上验证了方法的优越性。英文摘要：In this work, we study face verification in datasets where images of the sameindividuals exhibit significant age differences. This poses a major challengefor current face recognition and verification techniques. To address thisissue, we propose a novel approach that utilizes multitask learning and aWasserstein d