Pierre Abillama

Title: Pierre Abillama: Innovator in Neural Network Computation

Introduction

Pierre Abillama is a notable inventor based in Ann Arbor, MI (US). He has made significant contributions to the field of neural networks, particularly in developing efficient computational methods. His work is recognized for its potential to enhance the performance of deep learning systems.

Latest Patents

Abillama holds a patent for a "Hardware efficient weight structure for sparse deep neural networks." This innovative patent presents a computer-implemented method for performing computations with a neural network. The method involves several steps, including receiving a first input patch of data, applying a Walsh-Hadamard transform to yield a transformed input patch, and computing an element-wise product of the transformed input patch and a kernel of the neural network. The process continues with applying an inverse Walsh-Hadamard transform to produce an intermediate matrix and creating a first output patch from this matrix, which is smaller in size than the intermediate matrix. This patent showcases his expertise in optimizing neural network operations.

Career Highlights

Pierre Abillama is affiliated with the University of Michigan, where he contributes to research and development in advanced computational techniques. His work at the university allows him to collaborate with other experts in the field, further enhancing his contributions to technology and innovation.

Collaborations

One of his notable collaborators is Dennis Sylvester, with whom he has worked on various projects related to neural networks and computational efficiency.

Conclusion

Pierre Abillama's innovative work in neural network computation exemplifies the impact of research and development in technology. His contributions continue to influence the field and pave the way for future advancements.