Jinato Jiang

Title: Jinato Jiang: Innovator in Deep Neural Networks

Introduction

Jinato Jiang is a prominent inventor based in Great Falls, VA (US). He has made significant contributions to the field of artificial intelligence, particularly in the development of deep neural networks. His innovative approach has led to advancements in automatic speech recognition systems.

Latest Patents

Jinato Jiang holds a patent for "Centered, left- and right-shifted deep neural networks and their combinations." This patent focuses on the use of time-shifted deep neural networks (DNN) that are trained relative to one another. The method enhances recognition accuracy by combining these networks, which are based on an automatic speech recognition (ASR) system utilizing DNN and time-shifted features. The process involves training a regular ASR model to produce a first trained DNN, followed by fine-tuning additional layers with left- and right-shifted feature windows to create multiple DNNs. The resulting networks are then combined using techniques like ROVER to improve performance.

Career Highlights

Jinato Jiang is currently employed at Apptek, Inc., where he continues to innovate in the field of deep learning and speech recognition technologies. His work has been instrumental in pushing the boundaries of what is possible with neural networks.

Collaborations

Jinato has collaborated with notable colleagues such as Mudar Yaghi and Hassan Sawaf, contributing to a dynamic and innovative work environment.

Conclusion

Jinato Jiang's contributions to deep neural networks and automatic speech recognition systems exemplify the impact of innovative thinking in technology. His work continues to influence advancements in artificial intelligence and machine learning.