Moving Forward with Video Saliency: A New Dataset and Benchmark where Motion Matters
作者:Susmit Agrawal, Rebecca Wanner, Juliane Verwiebe, Matthias Tangemann, Matthias Bethge, Matthias Kümmerer · 单位:Tübingen AI Center, University of Tübingen, University of Toronto, Vector Institute · 会议/期刊:arXiv preprint · 方向:cs.CV · 发布日期:2026-10-05