A modular machine learning tool for holistic and fine-grained behavioral analysis.

Artificial intelligence techniques offer promising avenues for exploring human body features from videos, yet no freely accessible tool has reliably provided holistic and fine-grained behavioral analyses to date. To address this, we developed a machine learning tool based on a two-level approach: a first lower-level processing using computer vision for extracting fine-grained and comprehensive behavioral features such as skeleton or facial points, gaze, and action units; a second level of machine learning classification coupled with explainability providing modularity, to determine which behavioral features are triggered by specific environments. To validate our tool, we filmed 16 participants across six conditions, varying according to the presence of a person ("Pers"), a sound ("Snd"), or silence ("Rest"), and according to emotional levels using self-referential ("Self") and control ("Ctrl") stimuli.

Comparing the Auditory Distance and Externalization of Virtual Sound Sources Simulated Using Nonindividualized Stimuli.

When reproducing sounds over headphones, the simulated source can be externalized (i.e., perceived outside the head) or internalized (i.

PS1/gamma-secretase acts as rogue chaperone of glutamate transporter EAAT2/GLT-1 in Alzheimer's disease.

The recently discovered interaction between presenilin 1 (PS1), a subunit of γ-secretase involved in amyloid-β (Aβ) peptide production, and GLT-1, the major brain glutamate transporter (EAAT2 in the human), may link two pathological aspects of Alzheimer's disease: abnormal Aβ occurrence and neuronal network hyperactivity. In the current study, we employed a FRET-based fluorescence lifetime imaging microscopy (FLIM) to characterize the PS1/GLT-1 interaction in brain tissue from sporadic AD (sAD) patients. sAD brains showed significantly less PS1/GLT-1 interaction than those with frontotemporal lobar degeneration or non-demented controls.