MULTIDATA Summer Workshop 2025

Interested in the MULTIDATA project? Want to learn how to use our multimodal pipeline?

Join us for the MULTIDATA Summer Workshop on Tuesday, July 8, 2025, at the Max Planck Institute (Nijmegen), just before the ISGS conference.

This hands-on workshop will introduce participants to the MULTIDATA pipeline, covering transcription, speech analysis and OpenPose, followed by interactive sessions on applications in education and research. MULTIDATA is an ERASMUS PLUS KA220 HED project, involving the universities of Murcia, Radboud Nijmegen, and FAU Erlangen-Nürnberg as partners, and the MPI Psycholinguistics and the Red Hen Lab as associated partners. We offer an online platform for the study of multimodal communication (https://www.multi-data.eu/), including an AI-based pipeline to analyze speech and gesture data from videos, as well as other resources for developing audiovisual collections, as well as didactic materials and suggestions for exploiting video data inside and outside the classroom. This workshop will help you make the most of our open-source tools for analyzing gesture along with its co-occurring language and prosody.

Reserve your spot at https://multidata-summerworkshop-2025.eventbrite.com

📅 Date & Time: Tuesday, July 8, 2025 | 13:00 – 17:30

📍 Location: Max Planck Institute (Nijmegen)

📝 Registration: Free registration opens in April 2025

👥 Organizers: Cristóbal Pagán (University of Murcia), Peter Uhrig (FAU Erlangen-Nürnberg), Aslı Özyürek (MPI for Psycholinguistics)

MULTIDATA Summer Workshop Program

13:00–13:30 PM – Project Introduction

Speakers: Asli Özyürek, Cristóbal Pagán Cánovas, Peter Uhrig
Welcome from the MULTIDATA Project coordinators, followed by a general overview of the MULTIDATA Project.

13:30–13:40 PM – Platform Sign-Up

Speakers: Peter Uhrig, Raúl Sánchez Sánchez
How to access the MULTIDATA platform.

13:40–14:10 PM – Session + Demo 1: Extracting Aligned Transcriptions and Speech Metrics with MULTIDATA

Speakers: Raúl Sánchez Sánchez, Cristóbal Pagán Cánovas
This session demonstrates how to use the MULTIDATA platform to extract time-aligned transcriptions from video content.
Participants will learn how to:

- Generate ELAN-compatible files with a word-level alignment tier
- Obtain speech metrics (e.g., pitch and intensity) synchronized with video frame numbers

☕ COFFEE BREAK

14:35–15:25 PM – Session + Demo 2: Using Pose Estimation to Extract Gesture Features from Video Data

Speakers: Peter Uhrig, Brian Herreño Jiménez, Mounika Kanakanti, Ferdinand Paar
This session covers:

- Explanation of pose estimation as used for gesture annotation in MULTIDATA
- Data acquisition, cleaning, and processing workflows
- Demonstration of extracting kinematic features and using them in R dataframes

☕ COFFEE BREAK

15:50–16:20 PM – Showcase Session: Research Highlights Using the MULTIDATA Pipeline

Speakers: Mounika Kanakanti, Yassine Iabdounane, Armine Garibyan
Topics include:

- Applications of the MULTIDATA pipeline in multimodal communication research
- Spatial visualizations of gestures and signs (e.g., heatmaps)
- Exploratory analysis and insights using integrated tools

16:20–16:50 PM – Education Session: The MULTIDATA Pipeline Educational Resources and Applications

Speakers: Daniel Alcaraz Carrión, Sharice Clough
This session introduces:

- Current and upcoming online educational resources for multimodal communication
- In-class applications of MULTIDATA tools

16:50–17:20 PM – Practical Data Analysis Session: Exploiting MULTIDATA with Your Own Data

Speakers: Cristóbal Pagán Cánovas, Raúl Sánchez Sánchez
Hands-on session where participants can:

- Use MULTIDATA tools with their own video data
- Ask questions regarding computational resources and analysis
- Receive personalized support from the MULTIDATA team

17:20–17:30 PM – Workshop Closing

Speakers: Asli Özyürek, Cristóbal Pagán Cánovas, Peter Uhrig
Final remarks to close the workshop.

Avda. Teniente Flomesta, 5. 30003. Murcia
hello@multi-data.eu