Company Filing History:
Years Active: 2012
Title: Gil Dobry: Innovator in Audio Signal Classification
Introduction
Gil Dobry is an accomplished inventor based in Moshav Haniel, Israel. He has made significant contributions to the field of audio signal processing, particularly in the classification of audio signals. His innovative approach has led to the development of a unique method that enhances the understanding of audio interactions.
Latest Patents
Gil Dobry holds a patent for a "Method for speaker source classification." This method involves classifying a pair of audio signals into an agent audio signal and a customer audio signal. One embodiment of the patent focuses on unsupervised training, utilizing a training corpus that consists of multiple audio signal pairs. Each pair includes an agent signal and a customer signal, with the classification being unknown. The training process is based on the premise that agent signals are more similar to one another than customer signals. This leads to the determination of an agent cluster and a customer cluster, allowing for accurate classification of input signals. Another embodiment involves supervised training, where an agent model is generated to identify the agent signal based on higher scoring against the model.
Career Highlights
Gil Dobry has had a notable career, working with Nice Systems Ltd., a company recognized for its innovative solutions in customer engagement and analytics. His work has contributed to advancements in audio signal processing, showcasing his expertise and dedication to the field.
Collaborations
Throughout his career, Gil has collaborated with talented individuals such as Hila Lam and Moshe Wasserblat. These partnerships have fostered a creative environment that encourages innovation and the development of cutting-edge technologies.
Conclusion
Gil Dobry's contributions to audio signal classification exemplify the impact of innovative thinking in technology. His patent and work at Nice Systems Ltd. highlight his role as a key player in advancing audio processing methods.