Nanyang Ye

Title: Nanyang Ye and Deep Convolutional Neural Network Normalization.

Introduction

Nanyang Ye is an inventor based in Cambridge, Great Britain, whose patent work concerns machine learning and deep convolutional neural networks. His listed patent is assigned to Huawei Technologies Co., Ltd.

Latest Patents

Ye is associated with 1 patent, titled “Normalization in Deep Convolutional Neural Networks.” The invention describes a device with two neural network layers and an intervening normalization layer. During training, the normalization layer receives outputs for multiple samples in a batch, groups data values according to indices across two dimensions, and produces normalized outputs for the next layer.

The patent presents this approach as a way to support stable training across different batch sizes and generalization across multiple vision tasks, with the aim of improving training performance.

Career Highlights

The available information connects Ye’s patent activity with Huawei Technologies Co., Ltd. His listed invention focuses on normalization methods for training deep convolutional neural networks.

Collaborations

Ye’s listed coworkers include Xiaoyun Zhou and Jiacheng Sun. The available details do not specify their individual roles in the patent.

Conclusion

Nanyang Ye’s listed patent addresses neural network normalization, aiming to make deep convolutional network training more stable across batch sizes and useful across vision tasks.