Bumsoo Kang

Title: Innovations by Bumsoo Kang

Introduction

Bumsoo Kang is a notable inventor based in Austin, TX (US). He has made significant contributions to the field of artificial intelligence, holding a total of 4 patents. His work focuses on enhancing AI models through innovative methods that separate public and private knowledge.

Latest Patents

One of his latest patents is titled "Separating public and private knowledge in AI." This computer-implemented method links a private AI model to a public AI model, forming a combined AI model. The combined model is trained with private samples while keeping the public AI model fixed, ensuring that only the private AI model is trained with the private samples. Another significant patent is "Hamming distance based robust output encoding for improved generalization." This method generates a Hamming code-based target label for each class of a dataset, maximizing the Hamming distance between target labels. It trains a convolutional neural network with these Hamming codes to produce a trained AI model, addressing confusability between classes using a confusion matrix.

Career Highlights

Bumsoo Kang is currently employed at International Business Machines Corporation, commonly known as IBM. His work at IBM has allowed him to explore and develop advanced AI technologies that have practical applications in various industries.

Collaborations

Some of his notable coworkers include Minsik Cho and Christopher McCall Durham. Their collaborative efforts contribute to the innovative environment at IBM, fostering advancements in AI research and development.

Conclusion

Bumsoo Kang's contributions to artificial intelligence through his patents and work at IBM highlight his role as a leading inventor in the field. His innovative methods continue to push the boundaries of AI technology.