Researchers have introduced OpenVE-3M, a large-scale, open-source dataset designed for instruction-guided video editing. The dataset features two main categories of edits: spatially-aligned and non-spatially-aligned, with a focus on quality and diversity surpassing existing datasets. To evaluate performance, they also developed OpenVE-Bench, a benchmark comprising 431 video-edit pairs. A 5B parameter model, OpenVE-Edit, trained on this dataset, achieved state-of-the-art results on OpenVE-Bench, outperforming larger baseline models. AI
IMPACT This dataset and model advance the capabilities and evaluation of instruction-guided video editing, potentially improving tools for content creation.
RANK_REASON The cluster describes a new academic dataset and benchmark for video editing, along with a model trained on it. [lever_c_demoted from research: ic=1 ai=1.0]
- alphaXiv
- arXiv
- CatalyzeX
- DagsHub
- Gotit.pub
- Haoyang He
- Hugging Face
- OpenVE-3M
- OpenVE-Bench
- OpenVE-Edit
- ScienceCast
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
