arXiv:2510.26412v3 Announce Type: replace 
Abstract: Recent advances in text-to-video generation have achieved impressive performance on short clips, yet evaluating long-form generation under complex textual inputs remains a significant challenge. In response to this challenge, we present LoCoT2V-Bench, a benchmark for long video generation (LVG) featuring multi-scene prompts with hierarchical metadata (e.g., character settings and camera behaviors), constructed from collected real-world videos. We further propose LoCoT2V-Eval, a multi-dimensional framework covering perceptual quality, text-video alignment, temporal quality, dynamic quality, and Human Expectation Realization Degree (HERD), with an emphasis on aspects such as fine-grained text-video alignment and temporal character consistency. Experiments on 17 representative LVG models reveal pronounced capability disparities across evaluation dimensions, with strong perceptual quality and background consistency but markedly weaker fine-grained text-video alignment and character consistency. These findings suggest that improving prompt faithfulness and identity preservation remains a key challenge for long-form video generation. Our code and data are released at https://github.com/XqZeppelinhead0702/LoCoT2V-Bench

أدت التقدمات الأخيرة في توليد الفيديو من النص إلى تقديم LoCoT2V-Bench، وهو معيار مصمم لتقييم توليد الفيديو الطويل (LVG) باستخدام مدخلات نصية معقدة. يتضمن هذا المعيار مشاهد متعددة وبيانات وصفية هرمية، مما يعالج التحديات المرتبطة بتقييم مخرجات الفيديو الطويل، كما توضح التجارب على 17 نموذجًا من LVG التي تكشف عن تفاوتات ملحوظة في الأداء عبر أبعاد تقييم مختلفة.

Los recientes avances en la generación de videos a partir de texto han llevado a la introducción de LoCoT2V-Bench, un estándar diseñado para evaluar la generación de videos largos (LVG) con entradas textuales complejas. Este estándar incluye indicaciones de múltiples escenas y metadatos jerárquicos, abordando los desafíos de evaluar salidas de video de formato largo, como lo demuestran los experimentos en 17 modelos de LVG que revelan disparidades significativas en el rendimiento a través de diversas dimensiones de evaluación.

Des avancées récentes dans la génération de vidéos à partir de texte ont conduit à l'introduction de LoCoT2V-Bench, une référence conçue pour évaluer la génération de vidéos longues (LVG) avec des entrées textuelles complexes. Cette référence comprend des invites multi-scènes et des métadonnées hiérarchiques, répondant aux défis d'évaluation des sorties vidéo longues, comme le montrent des expériences sur 17 modèles LVG qui révèlent des disparités significatives de performance à travers diverses dimensions d'évaluation.

Recent advancements in text-to-video generation have led to the introduction of LoCoT2V-Bench, a benchmark designed for evaluating long video generation (LVG) with complex textual inputs. This benchmark includes multi-scene prompts and hierarchical metadata, addressing the challenges of assessing long-form video outputs, as highlighted by experiments on 17 LVG models that reveal significant disparities in performance across various evaluation dimensions.

LoCoT2V-Bench: Benchmarking Long-Form and Complex Text-to-Video Generation

One More Thing in AI – Your Shortcut to AI Mastery

LoCoT2V-Bench: Benchmarking Long-Form and Complex Text-to-Video Generation

Was this article worth reading? Share it

One More Thing in AI

vidvoi

VidBoard AI

Focal

Republiclabs.ai

Videotok

Ready to build your own newsroom?