- Hitoshi IYATOMI
- Applied Informatics
- Intelligent Information Processing Laboratory (IIPL)
- Research Fields
- Machine Learning, Computer Vision, Medical Engineering, Natural Language Processing
- Machine Learning, Deep Learning, Computer-Aided Diagnosis, Natural Language Processing, Dimensional Reduction
Our research interest lies on machine learning and its practical applications. We are investigating wide variety of research with our collaborators and currently there are four research groups in our lab; (a) plant, (b) medical, (c) natural language processing (NLP), and (d) high-dimensional data groups.
The plant group (a) is developing an image-based automated diagnosis system for crops as a part of national research project. The medical group (b) is currently focusing on skin and gastric cancers and develop an automated diagnosis support system. They are so-called “fine-grained” not easy computer-vision task. The motivation of those two groups is to establish low-cost and highly accurate explainable diagnosis system to find symptoms in their early stage. The NLP group (c) handles several NLP and signal processing studies including document classification, estimation of media bias, security application etc. The high-dimensional data group (d) aims to realize a content-based image retrieval on 3D brain MRI images. This groups also investigate related essential techniques on this, such as dimensional reduction, clustering, segmentation etc.
We are also investigating fundamental machine learning studies to achieve individual goal. They are (1) fight with “over-fitting”, one of the most intractable problems of deep learning and (2) realizing “explainable-AI”. All groups work effectively by sharing their progress and knowledge.