arXiv:2603. 13402v3 Announce Type: replace-cross Abstract: Current text-to-video models can make individual frames look convincing while still getting simple interactions wrong: objects move before contact, an intended action is skipped, a placed object keeps drifting, or a support relation breaks.
arXiv:2607. 24560v1 Announce Type: cross Abstract: We introduce EgoPlay, an event-triggered video-to-video editor for egocentric streams, obtained by fine-tuning a pretrained V2V diffusion transformer on event-conditioned data built primarily from Ego4D.
arXiv:2610.01039v1 Announce Type: new Abstract: While recent video generative models can synthesize high-fidelity videos, they struggle to portray plausible physical interactions and the resulting st...
arXiv:2609.38839v1 Announce Type: new Abstract: Long-horizon video generation requires models to effectively leverage an increasingly long generation history. As the generated history grows, retainin...
We introduce EgoPlay, an event-triggered video-to-video editor for egocentric streams, obtained by fine-tuning a pretrained V2V diffusion transformer on event-conditioned data built primarily from Ego4D. Given a monocular video and an event-triggered prompt of the form "when X happens, do Y," EgoPlay infers whether and when event X occurs, preserves pre-event frames, and applies edit Y only to the post-event continuation.
LiveProBench evaluates streaming video models on their ability to interact proactively, assessing whether they respond at appropriate times without explicit cues. The benchmark tests models at one‑second intervals across six subtasks that vary trigger ambiguity and timing tolerance, measuring response accuracy, silence rates, and duplicate responses. Results show that many models issue premature responses more often than missed ones, highlighting a significant shortfall in human‑like temporal decision making.
ProactiveBench evaluates streaming video models on their ability to interact proactively, rather than reactively. It tests models at one‑second intervals without explicit cues, using six subtasks that vary trigger ambiguity, timing tolerance, and response patterns. The benchmark measures both response and silence rates, distinguishing early, in‑window, and missed responses, and penalizes omissions and repetitions.
arXiv:2606. 18586v1 Announce Type: cross Abstract: Physical events are not understood by their names alone, but by the causal state changes that compose them.
The paper introduces VT-Contrast, a representation-level temporal counterfactual objective designed to improve temporal understanding in Video Language Models (VideoLMs). By supervising late-layer last-frame video tokens and contrasting order-preserving views with reordered counterfactuals graded by Kendall tau distance, VT-Contrast addresses the mismatch between ordered video input and text-based supervision. The method requires no architectural changes, is compatible with various VideoLM training tasks, and demonstrates improved performance on temporal understanding benchmarks.
arXiv:2602.08277v3 Announce Type: replace-cross Abstract: The landscape of AI video generation is undergoing a pivotal shift: moving beyond general generation - which relies on exhaustive prompt-engi...
arXiv:2609.40219v1 Announce Type: cross Abstract: World Action Models (WAMs) couple visual dynamics prediction with action generation, yet they do not explicitly support the reuse of action experienc...
OmniAssistBench is a new benchmark for evaluating omni-modal large language models (Omni-LLMs) as real‑time video assistants that actively guide users toward goals. The benchmark addresses the challenge of dynamic interaction paths by providing models with predefined priors from source videos, forcing them to follow the same routes as users. The dataset was constructed by reverse‑engineering existing Internet videos into multi‑turn clips, a process that required over 1,000 expert person‑hours. Results show that proprietary Gemini‑3‑Pro scores 66.4/100 while open‑source Qwen3‑Omni‑Instruct scores 51.2, revealing that current models often give incorrect or incomplete answers, struggle with visual prompts, and fail to maintain context or delay responses until target events.
arXiv:2608. 11605v1 Announce Type: new Abstract: World Action Models (WAMs) couple future visual prediction with robot action generation, enabling policies to model how the physical world evolves during interaction.