Thu PM1.L4.3
TEXT-TO-TEXT PRE-TRAINING WITH PARAPHRASING FOR IMPROVING TRANSFORMER-BASED IMAGE CAPTIONING
Ryo Masumura, Naoki Makishima, Mana Ihori, Akihiko Takashima, Tomohiro Tanaka, Shota Orihashi, NTT Corporation, Japan
Session:
Thu PM1.L4: Scene understanding Lecture
Track:
VIP - Visual Information Processing
Location:
Nautica
Presentation Time:
Thu, 7 Sep, 14:40 - 15:00 Finland Time (UTC +3)
Presentation
Discussion
Resources
No resources available.
Session Thu PM1.L4
Thu PM1.L4.1: Pose-aware Disentangled Multiscale Transformer for Pose Guided Person Image Generation
Kei Shibasaki, Masaaki Ikehara, Keio University, Japan
Thu PM1.L4.2: ACTIVE FACE RECOGNITION THROUGH VIEW SYNTHESIS
Efstratios Kakaletsis, Nikos Nikolaidis, Aristotle University of Thessaloniki, Greece
Thu PM1.L4.3: TEXT-TO-TEXT PRE-TRAINING WITH PARAPHRASING FOR IMPROVING TRANSFORMER-BASED IMAGE CAPTIONING
Ryo Masumura, Naoki Makishima, Mana Ihori, Akihiko Takashima, Tomohiro Tanaka, Shota Orihashi, NTT Corporation, Japan
Thu PM1.L4.4: A SCENE-ADAPTIVE FRAMEWORK FOR POSE-ORIENTED ABNORMAL EVENT DETECTION
Yuxing Yang, Newcastle University, United Kingdom; Zeyu Fu, University of Exeter, United Kingdom; Naqvi Syed Mohsen, Newcastle University, United Kingdom
Thu PM1.L4.5: GENERATING SALIENT SCENE GRAPHS WITH WEAK LANGUAGE SUPERVISION
Alexandros Benetatos, National Technical University of Athens, Greece; Markos Diomataris, Vassilis Pitsikalis, deeplab.ai, Greece; Petros Maragos, National Technical University of Athens, Greece