One of the lecturers at the Faculty of Computer Science (FILKOM) Brawijaya University (UB), namely Dr. Novanto Yudistira carried out a series of activities in Japan. Activities were carried out at The Institute of Scientific and Industrial Research, Osaka University, Suita, Osaka. Visits to Yagi, Matsushita and Nagahara Laboratories were carried out from 22/4 to 26/4, and became guest lecturers on 23/4 at 13.00 local time.
The topic presented was related to “Explainable Deep Learning for Multipurpose from Weakly Object Segmentation to Identifying Factors”. The material includes deep learning, which is not only related to training and inference, but the rich features in it can be used as explanations or to explain why deep learning classifies objects as cats or dogs, for example.
“The explanation shows that deep learning classifies the objects as dogs and cats based on their attributes such as body parts or eyes, and so on. This shows that deep learning behaves like humans when making decisions. If we succeed in carrying out such an explanation, we can use this explanation for other tasks, for example to segment objects without knowledge of the object’s precision, even to answer what factors cause COVID-19 cases to rise or fall, for example,” explained the man who is familiarly called Yudis.
Another activity while in Japan was carrying out joint publications with Prof. Yasushi Yagi, who is an expert in the field of computer vision and gait recognition. No less important, on that occasion an MoA was also carried out between FILKOM UB and The Institute of Scientific and Industrial Research, Osaka University, especially Yagi laboratory in terms of research collaboration, academic exchange, student exchange, and sharing ICT lab. [yudis/pon/UB PR/ Trans. Iir]
