Shang-Hong Lai

Title: Innovations of Shang-Hong Lai

Introduction

Shang-Hong Lai is a prominent inventor based in Hsinchu, Taiwan. He has made significant contributions to the field of image processing and deep learning, holding a total of 12 patents. His work focuses on developing advanced methods for image completion and structural estimation in indoor scenes.

Latest Patents

One of his latest patents is a method of image completion. This method involves constructing an image repair model and a series of conditional generative adversarial networks tailored to various object types. The process includes inputting training images for these networks to conduct corruption feature training. The method allows for the input of images needing repair, which are processed through the networks to generate multiple repaired images. A probability analyzer then assesses these images to select the most suitable one for output.

Another notable patent is the indoor scene structural estimation system based on a deep learning network. This system comprises a 2D encoder, several decoders for planes, edges, and corners, and a 3D encoder. The 2D encoder processes input images, while the decoders generate layout images that help in estimating the structure of indoor scenes.

Career Highlights

Shang-Hong Lai has worked with esteemed institutions such as Tsinghua University and Novatek Microelectronics Corporation. His experience in these organizations has allowed him to collaborate on various innovative projects that push the boundaries of technology.

Collaborations

Some of his notable coworkers include Chia-Te Liao and Tung-Ying Lee. Their collaborative efforts have contributed to the advancement of their respective fields.

Conclusion

Shang-Hong Lai's innovative work in image processing and deep learning has led to significant advancements in technology. His contributions are recognized through his numerous patents, showcasing his expertise and dedication to innovation.