keywords CONVOLUTIONAL NEURAL NETWORKS, DEEP LEARNING, GPUS
Reference persons SARAH AZIMI, LUCA STERPONE
Research Groups DAUIN - GR-05 - ELECTRONIC CAD & RELIABILITY GROUP - CAD
Thesis type RESEARCH
Description The "Signal for Help" is a signal-handed gesture that was created by the Canadian Women's Foundation to be used by individuals to alert others that they feel threatened and need help. The idea proposed in this thesis is to provide automatic detection of "Signal for Help", exploiting the video surveillance scenarios.
The key idea of this thesis is to train the CNN model for hand gesture recognition to detect and classify "Signal for Help" gestures. The model will be implemented on the state-of-the-art NVIDIA GPU (Jetson Nano).
Required skills Good knowledge of Python is welcome.
Deadline 01/04/2023 PROPONI LA TUA CANDIDATURA