作者位次 #3/3CVPR 2026Poster· Computer Vision & Pattern Recognition
TempoControl: Temporal Attention Guidance for Text-to-Video Models
Shira Schiber · Ofir Lindenbaum · Idan Schwartz
多模态OpenReview 项目页
首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。
实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。
Shira Schiber · Ofir Lindenbaum · Idan Schwartz
Guy Yariv · Idan Schwartz · Yossi Adi · Sagie Benaim
Commonsense reasoning often requires both textual and visual knowledge, yet Large Language Models (LLMs) trained solely on text lack visual grounding (e.g., “what color is an emperor penguin’s belly?”). Visual Language Models (VLMs) perform better on visually grounded tasks but face two limitatio...