The mission of the Section of Brain Function Information is to support collaborative studies using high field magnetic resonance imaging （3T and 7T） and to promote research on functional-anatomical mapping of the primate brain. We are actively promote collaborative studies ranging from basic research and development of MRI to clinical applications as well as studies on establishing standards for MRI procedures, including safety, applications, and quantitative analyses of the images. We are now trying to develop an algorithm to quantitatively and statistically handle image data of the brain generated by MRI. In addition to collaborative research, training junior researchers in MRI applications and basic neuroscience research are promoted.
Recently, we focus on the combination of functional MRI and deep learning. Specifically, we built artificial intelligence that could predict the price of art. We further apply individual optimization to the AI, which results in making the AI mimic individual’s preferences (Publicly offered research group in “Correspondence and fusion of artificial intelligence and brain science”). In our lab, students can learn how to analyze functional MRI data as well as how to use deep learning.
A schematic figure of transfer learning for vision-value converter, based on VGG 16. We used the same structure for vision-to-category transformation as VGG16 and added new layers for category-to-value transformation.
Introduce a researcher of NIPS.