Suji Kim

Title: Suji Kim: Innovator in Neural Codec and Image Segmentation Technologies

Introduction

Suji Kim is a prominent inventor based in Yongin-si, South Korea. She has made significant contributions to the fields of neural codecs and image segmentation. With a total of 3 patents to her name, her work continues to influence technological advancements in these areas.

Latest Patents

Suji Kim's latest patents include a "Method and apparatus with neural codec." This innovative neural codec features a first simulated predictor that predicts a first block corresponding to a target block within a current frame. The prediction is based on the target block of the to-be-predicted current frame and pixels of neighboring blocks. Additionally, it includes a second simulated predictor that predicts a second block using a reference block determined by a prediction mode. A selection network is also integrated to choose between the first and second blocks based on the prediction mode.

Another notable patent is the "Image segmentation method and device." This electronic device extracts feature data from an input image and calculates class maps using a classifier layer. It also generates cluster maps through a clustering layer, ultimately producing image segmentation data from these maps.

Career Highlights

Suji Kim is currently employed at Samsung Electronics Co., Ltd., where she continues to develop cutting-edge technologies. Her work has garnered attention for its innovative approach to complex problems in image processing and data compression.

Collaborations

Suji collaborates with talented coworkers, including Seungjun Shin and Dokwan Oh. Their combined expertise contributes to the success of their projects and the advancement of technology at Samsung.

Conclusion

Suji Kim's contributions to neural codecs and image segmentation highlight her role as a leading inventor in her field. Her innovative patents and collaborations with skilled professionals underscore her impact on technology and research.