Project Details

Multimodal deep learning for People Detection and Demographic Analytics

Dates: 2023
Principal Investigator: Dr. Sadam Al-Azani
Description: Detecting persons and analyzing their demographic characteristics (such as gender, age group, dialect, race/ethnicity, nationality, etc.) from multimodal video/audio data is very important and has several interesting applications. It is an important tool for AI conversational systems or human-computer interaction systems as a pre-processing step to understand people communities, and demographics. Consequently, those systems can adapt themselves to interact with the user according to his/her detected age, gender, race, dialect, emotions, among others. It is also an excellent opportunity for large companies to capitalize on, by extracting users’ profiles, sentiments, suggestions, and complaints on their products from reviews while considering the demographic characteristics. Consequently, they can improve and enhance their products/services to meet the needs of their customers. Demographic analysis can also be useful to overcome the real-world gender, age, and racial biases of AI-based detection and recognition systems.