A project team of the MEXT Grant-in-Aid for Scientific Research on Innovative Areas "Integrative Research toward Elucidation of Generative Brain Systems for Individuality," which is headed by Professor Tomohiro Shibata of the Department of Human Intelligence and Systems Engineering, has been developing a Markerless Representational Recording and Mining System to discover "individuality", and has recently developed the world's first system to estimate the motion of macaque monkeys in the natural environment from 2D images. The model of DeepLabCut, which can be used to create a model and preprint of the model, is now available. They have created a large open dataset of feature labels for monkeys in natural scenes, which can be used to train and test deep learning models for markerless motion capture in macaque monkeys. This dataset will help train and test deep learning models for markerless motion capture in macaque monkeys, as well as develop algorithms to accelerate the development of innovative behavioral analysis techniques for non-human primates using artificial intelligence in many fields, including neuroscience, medicine, and ecology. It is expected to contribute to research.

●Scientific Research on Innovative Areas FY2016-2020: Brain and Individuality

Example of performance evaluation for motion capture (feature position estimation)