Company Filing History:
Years Active: 2014-2021
Title: Innovations of Luhong Liang in Neural Network Optimization
Introduction
Luhong Liang is a prominent inventor based in Hong Kong, China, known for his significant contributions to the field of artificial intelligence and neural networks. With a total of five patents to his name, Liang has developed innovative methods that enhance the efficiency and cost-effectiveness of neural network implementations.
Latest Patents
Liang's latest patents include a "Method and system for bit-depth reduction in artificial neural networks" and "Residual quantization of bit-shift weights in an artificial neural network." The first patent introduces a bit-depth optimization engine that reduces hardware costs associated with neural networks. By generating costs for weights near bit-depth steps, the method allows for the selection of weights that can be optimized to create a low-bit-depth neural network, ultimately leading to lower manufacturing costs on Application-Specific Integrated Circuits (ASICs).
The second patent focuses on a neural network accelerator that reads encoded weights from memory. This innovation discards all 1 bits in a weight except for the first three, which are encoded as bit-shift values. This approach significantly reduces node complexity by minimizing the number of shifts required, allowing for efficient implementation using Look-Up Tables (LUTs) without the need for a Multiply-Accumulate (MAC) cell in Field-Programmable Gate Arrays (FPGAs).
Career Highlights
Luhong Liang is affiliated with the Hong Kong Applied Science and Technology Research Institute Company, Limited, where he continues to push the boundaries of research in artificial intelligence. His work has garnered attention for its practical applications in reducing costs and improving the performance of neural networks.
Collaborations
Liang collaborates with notable colleagues, including King Hung Chiu and Xuejiao Liu, contributing to a dynamic research environment that fosters innovation and technological advancement.
Conclusion
Luhong Liang's contributions to neural network optimization exemplify the intersection of innovation and practicality in technology. His patents not only advance the field of artificial intelligence but also pave the way for more efficient and cost-effective solutions in hardware implementation.