Cinema features: Cinema features such as the shot length, the scale of a shot, camera angle, camera level, and camera movements are among the main stylistic and narrative functions of movie, conveying meaning and inducing the viewer’s emotional state. We propose to build a DL model for estimating all the cinematographic features from movie frames, without the need to recover the 3D structure of the scene [python and machine learning].

Narrative experiences in movies: We want to investigate the effect of camera angle/level on viewers’ narrative experiences in the context of character engagement, emotional engagement, narrative presence, moral engagement with characters, and narrative engagement in general [basic statistics] [in collaboration with University of Glasgow and Vrije Universiteit Amsterdam].

Children's understanding of cartoons: Children are exposed to audiovisual content from birth yet our understanding of how children learn to comprehend audiovisual narratives is highly limited. The project aims to understand what is going on in the mind of children whilst they watch cartoons [python, basic statistics] [in collaboration with Birkbeck University (UK)].

All following activities are in collaboration with University of Glasgow (UK):

Alzheimer’s Disease trajectories: Running LOD-Brain on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and exploring relationships between brain structure volumes and cognitive decline [some coding and basic statistics]

Alzheimer’s Disease Classification: Build a classification model to distinguish between healthy individuals and those with Alzheimer’s disease based on MRI scans, potentially identifying early markers of the disease [python and machine learning]

Brain Age Prediction: Train a machine learning model to predict an individual’s age based on their brain structure, leveraging MRI data and exploring relationships between brain ageing and cognitive decline [python and machine learning]

Brain Age Prediction v2: Using a pre-trained DL MRI model for age estimation, explore relationships between predicted age and chronological age and various demographic or clinical factors [some coding and basic statistics]

Cortical Thickness analysis: We have a DL model for estimating CT, and we want to explore relationships between CT and ageing in healthy and clinical populations [some coding and basic statistics]

Apparent Motion: Using an apparent motion paradigm, we want to replicate the existence and spatiotemporal specificity of the motion masking effect in a digital setting (i.e., online experiment) [basic statistics]

MRI brain atlases: Investigate differences between atlases and retinotopic mapping in the visual cortex localisation in native space [python, brain imaging]

All following activities are in collaboration with Yonder s.r.l.:

Topic discovery in customers' feedback: We propose to work on techniques for customers’ reviews analysis, and in particular Aspect Term Extraction (ATE), by designing a pipeline made of tools, mostly based on Large Language Models (LLMs) [python and machine learning].