AMILab mainly focuses on the following research agenda: * Machine perception understanding - Multi-modal learning - Video (visual and temporal) understanding - 3D perception understanding * Data efficient learning - Prior based learning and inference - Minimal-supervised learning - Learning with human-in-the-loop
If you're curious about the detailed research areas being pursued at AMI Lab, please refer to the following introductory materials.
We believe that, in order to achieve Artificial Generic Intelligence, Sensing and Efficient Learning are the most important and fundamental blocks needed to be done first.
We are working toward developing machine perception capability by understanding human perception capability; thereby, intelligent agents can efficiently learn to understand about the real-world.
Besides this, AMILab members have broad research interests in Computer Vision & Machine Learning.
In particular, we put a high priority on the research that connects human and machine, i.e., human centric research.
Lastly, we are interested in the following general topics as well. * Deep neural networks * Invariant representation learning * Low-rank & sparse structure, and related optimization techniques including compressed sensing