Marco Cristani, Francesco Setti
- Università di Verona - Dip. Informatica
Tuesday, July 6, 2021
The ICE lab setup and the OPERA4.0 Cariverona project give the opportunity to explore the last computer vision and machine learning trends in multi-person activity analysis, and to apply some research findings which have been developed in our department. The talk will consider and discuss the state of art of classical yet critical monitoring operations as person detection and tracking, together with modern challenges as trajectory forecasting, re-identification and group detection, concluding with very recent problems as DPI detection and action prediction. For each topic, there will be a short description, why it is important for the Opera4.0 project and in general for the ICE Lab, a short review of the main techniques in the related literature and that we have recently published, together with the problems that are still open.