Wei-Hsiang Shen

Title: Innovations by Wei-Hsiang Shen in Neural Network Systems

Introduction

Wei-Hsiang Shen is a notable inventor based in Hsinchu, Taiwan. He has made significant contributions to the field of neural networks and signal processing. His innovative work has led to the development of a unique patent that enhances the capabilities of neural network systems.

Latest Patents

Wei-Hsiang Shen holds a patent for a "Neural network system and signal processing method." This invention includes a neural network system that comprises at least one processing unit and a neural network module. The signal processing method involves inputting a neural network input to the neural network module, generating an input at a previous layer of each convolutional transformer layer. The process includes performing pointwise convolution on the input, outputting a key tensor, and performing convolution on the input to output a value tensor. The attention embedding layer then processes the cascading tensor to output an attention tensor, which is used to generate an output tensor.

Career Highlights

Wei-Hsiang Shen is currently employed at Realtek Semiconductor Inc., where he continues to innovate in the field of signal processing and neural networks. His work is characterized by a strong focus on enhancing the efficiency and effectiveness of neural network systems.

Collaborations

Wei-Hsiang Shen collaborates with Chien-Hao Chen, contributing to advancements in their shared field of expertise.

Conclusion

Wei-Hsiang Shen's contributions to neural network systems and signal processing exemplify the innovative spirit of modern technology. His patent reflects a significant advancement in the field, showcasing his expertise and dedication to innovation.