In this study, we developed a technology to realize an instructor support system for group work by applying sound information processing technology. (1) Wearable devices worn by learners were improved by evaluating sound source separation features. (2) Deep learning identification algorithms were developed to visualize the participation attitudes of learners. (3) We developed a group work logging system and a support system for annotating group work participation information. (4) We developed a method for speaker anonymization of recorded group work speech by applying deep learning voice transformation. Due to the impact of the new coronavirus, we had to change our original plan and decided not to continue the face-to-face experiments, but we were able to obtain new knowledge that is useful for online education.