ListenToTheStadium
Analysis of the feelings of those attending a football match based on the relationship between the ambient sound of the match and what happens on the pitch, with the possibility of transferring it to other sports.
The ‘ListenToTheStadium’ project aims to develop software for analysing the sentiment of fans attending football matches in stadiums. Using the ambient audio captured during matches and combining it with a detailed archive of events on the pitch (passes, fouls, goals, dribbles, etc.), this software seeks to understand and quantify the reactions of spectators in relation to each moment and event of the match. The ultimate goal is to detect and analyse the mood of the spectators, relating it to the actions of the team and the players, both own and opponents.
The objectives of the project are divided into two phases:
- Software Development: Design and implement a software system that integrates advanced audio processing and sentiment analysis techniques. Develop algorithms capable of identifying and quantifying spectator sentiment from ambient audio.
- Data Mining: Audio data captured during matches will be extracted and analysed. This data will be integrated with an archive containing detailed information of all events on the pitch.
Regarding the methodology, audio pre-processing techniques will be implemented to filter out unwanted noise and improve the quality of the signal. Signal processing algorithms will be developed for the detection of audio patterns associated with different feelings (elation, disappointment, anger, etc.). Machine learning and sentiment analysis techniques will be used to classify and quantify the emotional reactions of viewers. A synchronisation system will be developed between the audio data and the match event archive, ensuring that each spectator reaction is correctly related to the corresponding event.
- Complexity:9/10
- CATEGORY 2024-2025 , Cantabria Univ.
- TAGS Python , Pytorch





