超球面流形上神经网络学习的改进
Improvising the Learning of Neural Networks on Hyperspherical Manifold
摘要：
卷积神经网络（CNN）在监督设置中的影响提供了巨大的性能提升。从超球面流形上操作的CNN学习的表示在人脸识别、人脸识别和其他监督任务中产生了有见地的结果。广泛的激活函数开发与超球直觉，其性能优于softmax在欧几里得空间。这项研究的主要目的是提供见解。首先，利用赤平投影将数据从欧氏空间（$\mathbb{R}^{n}$）转换到超球面流形（$\mathbb{S}^{n}$），分析了角度裕度损失的性能。其次，从理论和实践上证明了利用赤平投影在超球面上构造决策边界，为神经网络的学习提供了必要条件。实验已经证明，对现有技术水平的角裕度目标函数应用赤平投影改进了标准图像分类数据集（CIFAR-10，100）的性能。此外，我们在疟疾薄血涂片图像上进行了实验，得到了有效的结果。该代码可在https://github.com/barulalithb/stereo-angular-margin上公开获取。
英文摘要：
The impact of convolution neural networks (CNNs) in the supervised settings provided tremendous increment in performance. The representations learned from CNN's operated on hyperspherical manifold led to insightful outcomes in face recognition, face identification, and other supervised tasks. A broad range of activation functions were developed with hypersphere intuition which performs superior to softmax in euclidean space. The main motive of this research is to provide insights. First, the stereographic projection is implied to transform data from Euclidean space ($\mathbb{R}^{n}$) to hyperspherical manifold ($\mathbb{S}^{n}$) to analyze the performance of angular margin losses. Secondly, proving theoretically and practically that decision boundaries constructed on hypersphere using stereographic projection obliges the learning of neural networks. Experiments have demonstrated that applying stereographic projection on existing state-of-the-art angular margin objective functions improved performance for standard image classification data sets (CIFAR-10,100). Further, we ran our experiments on malaria-thin blood smear images, resulting in effective outcomes. The code is publicly available at:https://github.com/barulalithb/stereo-angular-margin.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
