Company Filing History:
Years Active: 2016-2021
Title: **The Innovations of Inventor Hanqing Lu**
Introduction
Hanqing Lu is an accomplished inventor based in Beijing, China, known for his significant contributions to the field of deep learning and image processing. With a total of three patents to his name, his work focuses on enhancing neural networks and image matching techniques, which has broad applications in technology and artificial intelligence.
Latest Patents
Among his latest patents, Hanqing Lu has developed a "Deep convolutional neural network acceleration and compression method based on parameter quantification." This innovative method involves quantizing parameters of deep convolutional neural networks (CNNs) to create subcode books and corresponding index values. By using these quantized parameters, the application enhances the acceleration and compression of CNNs, improving their efficiency and performance.
Another notable patent is the "Image matching method based on cascaded binary encoding." This method improves the speed and quality of image matching by utilizing a multi-table hashing look-up to filter candidate key-points and project them into a high-dimensional Hamming space. The resulting hashing table allows for optimal matching of key-points, making the technique suitable for efficient and accurate image matching.
Career Highlights
Hanqing Lu has had an impressive career, working with prestigious organizations such as the Chinese Academy of Sciences and Nokia Technologies Oy. His roles within these institutions have allowed him to push the boundaries of innovation and collaborate on cutting-edge research projects.
Collaborations
Throughout his career, Hanqing Lu has collaborated with esteemed colleagues, including Jian Cheng and Jiaxiang Wu. Their combined expertise and innovation have driven forward projects that contribute to advancements in their respective fields.
Conclusion
Hanqing Lu's contributions to the landscape of technology and artificial intelligence are commendable. His patented methods have the potential to propel deeper understanding and application of deep learning and image matching, which will undoubtedly pave the way for future innovations in various technological sectors.