Gaze Prediction from Egocentric Point of View
The project aim to build computer vision algorithms for the prediction of the gaze attention of a person wearing a camera (e.g. Smart Glasses). The algorithms will exploit Machine Learning techniques (e.g. Deep Learning) to make inference from sequences of videos. An experimental dataset will be built, labeled and exploited for evaluation purposes.
Required Competences:
Image processing, computer vision, machine learning, python/matlab programming.
Tutor: prof. Giovanni Maria Farinella.
Study cycle:
undergraduate
Languages skills required:
English B2
Length:
4 months
Period:
second semester
Summer traineeships:
no
Insurance:
Accident insurance during working hours only and Liability insurance
Benefits:
none
Traineeship type:
Erasmus traineeship