This article previews a talk by Gül Varol from Ecole des Ponts ParisTech on bridging natural language and 3D human motions. The talk will cover text-to-motion synthesis using generative models and text-to-motion retrieval models based on the ACTOR, TEMOS, TMR, TEACH, and SINC papers. Varol's research interests include video representation learning, human motion synthesis, and sign languages. Why it matters: Research in this area could enable more intuitive human-computer interaction and new applications in areas like virtual reality and robotics.
FancyVideo, a new video generator, introduces a Cross-frame Textual Guidance Module (CTGM) to enhance text-to-video models. CTGM uses a Temporal Information Injector and Temporal Affinity Refiner to achieve frame-specific textual guidance, improving comprehension of temporal logic. Experiments on the EvalCrafter benchmark demonstrate FancyVideo's state-of-the-art performance in generating dynamic and consistent videos, also supporting image-to-video tasks.
Nicu Sebe from the University of Trento presented recent work on video generation, focusing on animating objects in a source image using external information like labels, driving videos, or text. He introduced a Learnable Game Engine (LGE) trained from monocular annotated videos, which maintains states of scenes, objects, and agents to render controllable viewpoints. Why it matters: This talk highlights advancements in cross-modal AI, potentially enabling new applications in gaming, simulation, and content creation within the region.
Video-ChatGPT is a new multimodal model that combines a video-adapted visual encoder with a large language model (LLM) to enable detailed video understanding and conversation. The authors introduce a new dataset of 100,000 video-instruction pairs for training the model. They also develop a quantitative evaluation framework for video-based dialogue models.
Researchers from MBZUAI have introduced VideoMolmo, a large multimodal model for spatio-temporal pointing conditioned on textual descriptions. The model incorporates a temporal module with an attention mechanism and a temporal mask fusion pipeline using SAM2 for improved coherence across video sequences. They also curated a dataset of 72k video-caption pairs and introduced VPoS-Bench, a benchmark for evaluating generalization across real-world scenarios, with code and models publicly available.
A new approach to composed video retrieval (CoVR) is presented, which leverages large multimodal models to infer causal and temporal consequences implied by an edit. The method aligns reasoned queries to candidate videos without task-specific finetuning. A new benchmark, CoVR-Reason, is introduced to evaluate reasoning in CoVR.