Tesi etd-07072025-110929 |
Link copiato negli appunti
Tipo di tesi
Tesi di laurea magistrale
Autore
FOCACCI, EDOARDO
URN
etd-07072025-110929
Titolo
Facial recognition, development of an LLM-based short and Long-Term Memory for Social Robots
Dipartimento
INGEGNERIA DELL'INFORMAZIONE
Corso di studi
ARTIFICIAL INTELLIGENCE AND DATA ENGINEERING
Relatori
relatore Galatolo, Federico Andrea
correlatore Cominelli, Lorenzo
correlatore Cominelli, Lorenzo
Parole chiave
- Face detection
- Face recognition
Data inizio appello
23/07/2025
Consultabilità
Tesi non consultabile
Riassunto
Natural interaction between humans and social robots requires advanced recognition and memory capabilities to ensure continuity and coherence in conversations. This research aims to develop a facial recognition module for social robots, integrating a memory system based on Large Language Models (LLMs) to enhance both short- and long-term recall. The proposed system will enable the robot to identify individuals through facial features and store this information to recognize them within the same conversation session as well as over extended periods.
The study will begin with a review of state-of-the-art facial recognition techniques and memory models used in AI-driven conversational agents. The implementation phase will involve designing and developing a functional module that combines facial recognition with an LLM-based memory system. The goal is to allow the robot to associate recognized faces with contextual and conversational data, improving engagement and personalization in human-robot interactions.
The final evaluation will assess the system’s accuracy in facial recognition, the effectiveness of memory retention over time, and its impact on user experience. This research contributes to the development of more interactive and context-aware social robots, enhancing their ability to adapt and respond naturally in multi-session interactions.
The study will begin with a review of state-of-the-art facial recognition techniques and memory models used in AI-driven conversational agents. The implementation phase will involve designing and developing a functional module that combines facial recognition with an LLM-based memory system. The goal is to allow the robot to associate recognized faces with contextual and conversational data, improving engagement and personalization in human-robot interactions.
The final evaluation will assess the system’s accuracy in facial recognition, the effectiveness of memory retention over time, and its impact on user experience. This research contributes to the development of more interactive and context-aware social robots, enhancing their ability to adapt and respond naturally in multi-session interactions.
File
Nome file | Dimensione |
---|---|
Tesi non consultabile. |