Reuben Xin Hong Tan

Title: Reuben Xin Hong Tan: Innovator in Multi-Modal Localization

Introduction

Reuben Xin Hong Tan is a notable inventor based in San Jose, California. He has made significant contributions to the field of multi-modal localization, particularly through his innovative patent. His work focuses on enhancing the interaction between image data and corresponding text or audio data.

Latest Patents

Reuben holds a patent titled "Localization of narrations in image data." This patent outlines methods, systems, and computer storage media designed for multi-modal localization. The technology involves receiving input data that includes image data along with corresponding text or audio data. A phrase is extracted from the text or audio, and a neural network system is employed to spatially and temporally localize the phrase within the image data. The system features multiple cross-modal attention layers that compare features across different modalities without overlapping features of the same modality. This innovative approach allows for the identification of specific regions or subsets of pixels within the image data that correspond to the extracted phrase, ultimately presenting a localization indicator for display.

Career Highlights

Reuben is currently employed at Adobe Inc., where he continues to develop and refine his innovative ideas. His work at Adobe has positioned him as a key player in the advancement of technologies that integrate various forms of data.

Collaborations

Reuben collaborates with talented individuals such as Hailin Jin and Bryan Russell. Their combined expertise contributes to the success of their projects and the development of cutting-edge technologies.

Conclusion

Reuben Xin Hong Tan is a pioneering inventor whose work in multi-modal localization is shaping the future of data interaction. His contributions at Adobe Inc. and his innovative patent demonstrate his commitment to advancing technology in meaningful ways.