This inventor holds 3 USPTO granted patents. Top assignee: Kabushiki Kaisha Ubitus. Active years: 2022-2025.
Company Filing History:
Years Active: 2022-2025
Title: Innovations by Ying-Hau Wu
Introduction
Ying-Hau Wu is a notable inventor based in Taipei, Taiwan. He has made significant contributions to the field of artificial intelligence and media enhancement. With a total of 3 patents, his work focuses on improving the performance of AI systems in various applications.
Latest Patents
One of his latest patents is a method for training an AI bot in computer games. This invention discloses a pure end-to-end deep reinforcement learning approach for training a car racing game AI bot. It utilizes only the velocity information extracted from the screen for both training and testing phases, without relying on any internal state from the game environment, such as the car's facing angle. The AI bot developed through this method can outperform the average performance of human players. Additionally, the reward function is designed to consist solely of the velocity value, employing the Ape-X distributed training framework combined with a variant of Deep Q Network to address the sparse training signal problem caused by the original reward function design. Furthermore, a limit learner rate method is introduced to enhance training efficiency and performance. The AI bot trained in this manner achieves performance levels that are close to professional players.
Another significant patent by Ying-Hau Wu is a method for enhancing the quality of media. This method employs an artificial intelligence enhancing model built into the client device to improve the quality of video streams. The AI enhancement module is pre-trained using a neural network on the server to analyze the differences between decoded images and raw images generated by the server. The AI enhancement module enhances decoded images using algorithms defined by analyzing these differences, resulting in enhanced images that are visually more similar to the raw images than the decoded images.
Career Highlights
Ying-Hau Wu is currently employed at Kabushiki Kaisha Ubitus, where he continues to innovate in the field of AI and media technology. His work has garnered attention for its practical applications and effectiveness in enhancing user experiences.
Collaborations
Some of his notable coworkers include Chiu-Chou Lin and I-Chen Wu, who contribute to the collaborative efforts in their projects.
Conclusion
Ying-Hau Wu's contributions to the field of artificial intelligence and media enhancement demonstrate his innovative spirit and commitment to advancing technology. His patents reflect a deep understanding of AI applications and their potential to improve user experiences in gaming and media.