Hirofumi majored in applied chemistry as his undergraduate and studied influenza virus during his Master's degree. After 1 year of research in the field of regenerative medicine using iPS cells, he proceeded to his Ph.D. study where he developed ultra-high-speed imaging flow cytometry including image analysis pipelines using deep learning. He worked at Chan Zuckerberg Biohub initially as a fellow supported by JSPS (Japan Society for the Promotion of Science) and was later promoted to Scientist I, where he developed a deep-learning method to find out protein localization solely from fluorescence images. He is currently working at Pictor Labs to develop virtual staining techniques using deep learning. As suggested by his interdisciplinary background, however, his research interest, as well as expertise, spans a broad range of subjects from cell biology, and virology to optics, microfluidics, and deep learning.