MISTER is a project that aims to innovate and automate the mechanisms of Match Analysis in football. The basic idea is to improve video analysis and integrate it with that derived from other sources such as wearable devices or expert information.
Currently, data production is based on manual tag acquisition, through the use of sw that supports these operations: the detection takes place directly from video, acquired through special online platforms or from filming data is then produced by exporting the data directly; the data is then produced by exporting the data directly. tags, manually detected by a technician in performance analysis, in xls format: all data is aggregated and then exported; through the process of parsing of exported files, data processing takes place, an automated process and interactive in the development phase.
This project aims to improve productivity by accelerating activities of manual tagging without affecting the quality, exploiting:

1. Multi-source acquisition systems in order to increase the amount of available information;
2. Integration of voice/touch interfaces to speed up data tagging;
3. Machine learning algorithms applied in post processing to improve the quality of the proposed data according to semi-automated techniques that replace the current totally manual techniques.