Home  | Publications | LHC+25

When and Where Do Events Switch in Multi-Event Video Generation?

MCML Authors

Link to Profile Thomas Seidl PI Matchmaking

Thomas Seidl

Prof. Dr.

Director

Database Systems, Data Mining and AI

Link to Profile Daniel Cremers PI Matchmaking

Daniel Cremers

Prof. Dr.

Director

Computer Vision & Artificial Intelligence

Link to Profile Volker Tresp

Volker Tresp

Prof. Dr.

Principal Investigator

Database Systems, Data Mining and AI

Abstract

Text-to-video (T2V) generation has surged in response to challenging questions, especially when a long video must depict multiple sequential events with temporal coherence and controllable content. Existing methods that extend to multi-event generation omit an inspection of the intrinsic factor in event shifting. The paper aims to answer the central question: When and where multi-event prompts control event transition during T2V generation. This work introduces MEve, a self-curated prompt suite for evaluating multi-event text-to-video (T2V) generation, and conducts a systematic study of two representative model families, i.e., OpenSora and CogVideoX. Extensive experiments demonstrate the importance of early intervention in denoising steps and block-wise model layers, revealing the essential factor for multi-event video generation and highlighting the possibilities for multi-event conditioning in future models.

inproceedings

LongVid-Foundations @ICCV 2025

1st Workshop on Long Multi-Scene Video Foundations: Generation, Understanding and Evaluation at the IEEE/CVF International Conference on Computer Vision. Honolulu, Hawai'i, Oct 19-23, 2025. To be published. Preprint available.

Authors

R. Liao • G. Huang • Q. Cheng • T. Seidl • D. Cremers • V. Tresp

Links

Research Areas

 A3 | Computational Models

 B1 | Computer Vision

BibTeXKey: LHC+25

Back to Top