Yumin Suh

Title: Yumin Suh: Innovator in Facial Recognition and Semantic Segmentation

Introduction

Yumin Suh is a prominent inventor based in Santa Clara, California. He has made significant contributions to the fields of facial recognition and semantic segmentation, holding a total of three patents. His innovative work has the potential to impact various applications in technology and artificial intelligence.

Latest Patents

One of Yumin Suh's latest patents is the "Face-aware person re-identification system." This invention presents a method for employing facial information in unsupervised person re-identification. The method includes extracting body features from a first data stream and head features from a second data stream. It outputs a body descriptor vector and a head descriptor vector, which are then concatenated to enable a model to generate a comprehensive descriptor vector.

Another notable patent is "Multi-domain semantic segmentation with label shifts." This invention involves methods and systems for training a model by combining data from multiple datasets that have different respective label spaces. Relationships between labels in these different spaces are identified, and a unified neural network model is trained using the combined data and identified relationships. This results in a unified model that utilizes a class relational binary cross-entropy loss.

Career Highlights

Yumin Suh is currently employed at NEC Corporation, where he continues to develop innovative technologies. His work focuses on enhancing machine learning models and improving the accuracy of facial recognition systems. His contributions have been instrumental in advancing the capabilities of artificial intelligence in real-world applications.

Collaborations

Yumin has collaborated with talented individuals such as Yi-Hsuan Tsai and Masoud Faraki. These collaborations have fostered a creative environment that encourages the exchange of ideas and the development of cutting-edge technologies.

Conclusion

Yumin Suh is a distinguished inventor whose work in facial recognition and semantic segmentation is paving the way for future advancements in technology. His patents reflect a commitment to innovation and a deep understanding of complex systems.