Kunpeng Li

Title: Innovations by Kunpeng Li in Weakly-Supervised Semantic Segmentation

Introduction

Kunpeng Li, an accomplished inventor based in Somerville, MA, has made significant contributions to the field of computer vision. With a patent to his name, he has demonstrated expertise in developing advanced methodologies that enhance the effectiveness of image analysis.

Latest Patents

Kunpeng Li's patent, titled "Weakly-supervised semantic segmentation with self-guidance," focuses on systems, methods, and computer-readable media designed for performing weakly supervised semantic segmentation of input images. This innovative approach utilizes self-guidance on attention maps during training, enabling a Guided Attention Inference Network (GAIN) to holistically focus on objects in images rather than merely on the most discriminative parts. The patent cleverly incorporates both a classification loss function and an attention mining loss function to provide self-guidance, and it suggests that additional supervision can be derived from a selection of pixel-level labeled images to further improve the GAIN's semantic segmentation capabilities.

Career Highlights

Kunpeng Li is currently employed at Siemens Aktiengesellschaft, where he contributes his expertise in image processing and machine learning. His work is instrumental in developing cutting-edge technologies that benefit various industries, particularly in the realm of automation and intelligent systems.

Collaborations

Throughout his career, Kunpeng has collaborated with talented colleagues such as Ziyan Wu and Kuan-Chuan Peng. These partnerships have allowed him to leverage collective knowledge and skills, ultimately enhancing the innovations produced through their collective efforts.

Conclusion

Kunpeng Li exemplifies the spirit of innovation in the technology sector, particularly in computer vision. His patent reflects a deep understanding of weakly supervised techniques and highlights his contributions to advancements in semantic segmentation. Through his work at Siemens and collaborations with skilled coworkers, he continues to push the boundaries of what's possible in image analysis.