Skip to content

Program

Events

The Fall 2026 speaker series runs September through December, with seven talks. Times and format vary by session. Seats and RSVPs are managed on Luma; sessions already held move to the record below.

Subscribe to the series ↓Live feed. Sessions update on every site rebuild.
Next sessionFall 2026 · Sep 10 – Dec 3
01

Speaker series

Fall 2026 · Sep 101 / 7
Confirmed

Dr. Laurent Itti

Attention and Computational Vision (series kickoff)

Professor of computer science, psychology, and neuroscience at USC, where he directs the iLab. His work spans visual attention, computational neuroscience, and machine vision.

Sep 104:00 PM PTIn personReserve a seat ↗Add to calendar ↓

Arrow keys work too. The waveform above a confirmed name is a signature derived from the name. It is ornament, not data.

Confirmed talks are listed with speaker and topic. Slots marked Invited have a held date and a pending invitation; names and topics are published once the speaker confirms. Sessions already held move to the record below.

02

Reserve a seat

Seats and RSVPs are managed on Luma. Register for any session below, or follow the calendar to be notified when new sessions are posted.

Registration handled by Luma · luma.com/praxiscompsych ↗

03

Record

Held · Sep 10, 2026 · Speaker series · Kickoff

Attention and Computational Vision with Dr. Laurent Itti

Dr. Laurent Itti opened the Fall 2026 series with two decades of work on visual saliency and its route into clinical eye-tracking, including a 15-minute video that separates ADHD from Fetal Alcohol Spectrum Disorder at 77% accuracy. Recorded as episode 1 of the PRAXIS podcast.

Listen on Spotify

Held · Sep 1, 2026 · Live demo · Leavey Library, Room 202L

ClinicalWhisper, live on a real recording

A live demonstration of ClinicalWhisper, the speech model adapted for clinical interview audio, run in session on a real recording: transcription, speaker separation, acoustic features, and the score. Recorded; the recording is embedded below.

04

Interested in speaking?

If you apply computational methods to psychiatric questions (modeling, imaging, digital phenotyping, or clinical machine learning), we would like to host you. Talks run 45 minutes plus discussion, in person or over Zoom.