Saeed Mosayyebpour

Title: Innovations by Saeed Mosayyebpour

Introduction

Saeed Mosayyebpour is an accomplished inventor based in Irvine, California. He has made significant contributions to the field of neural networks and classification systems, holding a total of three patents. His work focuses on enhancing the efficiency and accuracy of keyword detection and classification processes.

Latest Patents

One of his latest patents is titled "Robust start-end point detection algorithm using neural network." This invention features an end detector that receives feature data to identify the endpoint of a keyword. Additionally, it includes a start detector that processes the feature data associated with input frames to detect the start point of the keyword. Both detectors utilize neural networks trained through specific cost functions to optimize performance.

Another notable patent is "Binary and multi-class classification systems and methods using one spike connectionist temporal classification." This invention comprises a neural network designed for classifying input data. It includes a training dataset with pre-segmented, labeled samples and a classification training module that trains the neural network. The module features both forward and backward pass processing to enhance classification accuracy.

Career Highlights

Saeed Mosayyebpour is currently employed at Synaptics Corporation, where he continues to innovate in the field of technology. His work has been instrumental in developing advanced algorithms that improve the functionality of neural networks.

Collaborations

He collaborates with talented individuals such as Trausti Thormundsson and Francesco Nesta, contributing to a dynamic work environment that fosters innovation and creativity.

Conclusion

Saeed Mosayyebpour's contributions to neural networks and classification systems demonstrate his expertise and commitment to advancing technology. His patents reflect a deep understanding of complex algorithms and their practical applications.