[visionlist] [CfP] - 1st CfP - MediaEval 2025 - Predicting Movie and Commercial Memorability

Mihai-Gabriel CONSTANTIN (115317) mihai.constantin84 at upb.ro
Tue May 27 15:06:48 -05 2025


*******************************************************
1st CALL FOR PARTICIPATION
MediaEval 2025 - Predicting Movie and Commercial Memorability
https://multimediaeval.github.io/editions/2025/tasks/memorability/
*******************************************************
Introduction
The Predicting Movie and Commercial Memorability at MediaEval 2025 invites researchers to develop systems that predict how likely a video is to be remembered by viewers after varying time intervals. This year introduces two new datasets: professionally produced movie excerpts and commercial videos, alongside EEG data from human viewers, enabling exploration of how narrative, emotional, and cognitive factors influence long-term memorability.
Participants will build models to predict both video and brand memorability, and explore EEG-based recall detection—offering new opportunities for interdisciplinary research in multimedia, machine learning, and neurocognition.
________________________________
Subtasks
Subtask 1: Movie Memorability

  *
Challenge 1.1 – How memorable is this video?
Predict long-term memorability scores for movie clips using video-based features.
  *
Challenge 1.2 – Is this person familiar with this video?
Use EEG data to detect whether a viewer remembers having seen a video clip before, without relying on any external information about the clip.

Subtask 2: Commercial/Ad Memorability

  *
Challenge 2.1 – How memorable is this commercial?
Predict the memorability score of commercial videos using visual and audio features.
  *
Challenge 2.2 – Can you predict the brand memorability?
Predict how well viewers remember the brand featured in a commercial video.

________________________________
Target Communities
We invite participants from a wide range of disciplines, including but not limited to:

  *
Multimedia analysis and video understanding
  *
Human cognition and memory research
  *
Affective and perceptual computing
  *
EEG and neurophysiological signal processing
  *
Machine learning and deep learning
  *
Human–computer interaction and user experience design
  *
Educational technology and content optimization
  *
Marketing science and brand research

All are welcome—no previous experience with EEG or memorability research is required.
________________________________
Data
Participants will receive two annotated datasets:

  *
Movie Memorability Dataset
Includes professionally produced movie excerpts with long-term memorability scores and EEG recordings from 27 participants (64 channels, 3484 epochs).
  *
Commercial Video Dataset
Includes 424 commercial videos and associated metadata with video and brand memorability scores.

Both datasets include pre-extracted audio/video features.
________________________________
Workshop
All participants are invited to submit a working notes paper and present their systems and results at the MediaEval 2025 Workshop, which will be held during 25–26 October 2025 in Dublin, Ireland. The workshop provides a supportive forum to exchange ideas and insights on computational models, cognitive theory, and practical applications.
________________________________
Important Dates (Tentative)
Milestone
Date
Registration opens
May 2025
Development data release
May 2025
Test data release
June 2025
Run submission deadline
24 September 2025
Working notes paper due
8 October 2025
MediaEval 2025 Workshop
25–26 October 2025, Dublin + Online
________________________________
Task Coordination

  *
Alba García Seco de Herrera – UNED, Spain (Task Lead)
  *
Sebastian Halder, Ana Matran-Fernandez – University of Essex, UK
  *
Mihai Gabriel Constantin, Bogdan Ionescu – University Politehnica of Bucharest, Romania
  *
Claire-Hélène Demarty – InterDigital, France
  *
Rukiye Savran Kiziltepe – Ankara University, Türkiye
  *
Iván Martín-Fernández, Manuel Gil Martín – Technical University of Madrid (UPM), Spain

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://visionscience.com/pipermail/visionlist_visionscience.com/attachments/20250527/e16915f5/attachment.html>


More information about the visionlist mailing list