Program / GS-06: Generative & Vision-Language Foundation Models

GS-06: Generative & Vision-Language Foundation Models

November 05 (Thursday) · 10:20–12:20 · Room C+D

Chair: Seiki Ubukata (Osaka Metropolitan University)

Back to timetable

10:20–10:40Paper ID: 24Full paper

Posture Queries in Open-Vocabulary Detection: Phrasing-Invariant Failure and Recovery

Yun A Kim (Sungkonghoe University); Sungjun Hong (Sungkonghoe University); Sangyun Lee (Sungkonghoe University)*

10:40–11:00Paper ID: 127Full paper

CountFix: A Generator-Agnostic Feedback-Control Framework for Object Counting in Text-to-Image Models

Hanqing Sun (The University of Electro-Communications)*; Tomonori Hashiyama (The University of Electro-Communications)

11:00–11:20Paper ID: 142Full paper

Impact of Visually Descriptive Text and Image Captions on CLIP-Based Cloud Classification: Per-Condition Modality Contribution Analysis

Taiga Nishino (University of the Ryukyus)*; Satoshi Endo (University of the Ryukyus); Hiroki Takei (Weathernews Inc.)

11:20–11:40Paper ID: 172Full paper

What Does Tangent Spectrum Regularization Do to a Diffusion Tokenizer? A Controlled Dissection

Eunha Seo (Sungkyunkwan University); Donghyuk Kang (Sungkyunkwan University); Jaekwang KIM (Sungkyunkwan University)*

11:40–12:00Paper ID: 219Full paper

Coherence-Weighted Future-Frame Prediction in Indoor Environments.

Arpan Ghosh (Sungkyunkwan University)*; Jinho Kim (Sungkyunkwan University); Ye-Chan An (Sungkyunkwan University); Tae-Yong Kuc (Sungkyunkwan University)

Presentations finish at 12:00. The scheduled session ends at 12:20.

Full paper: 15 minutes presentation + 5 minutes Q&A. Short, abstract-only and late breaking papers: 10 minutes presentation + 5 minutes Q&A. Speakers should bring their own laptop; presentation PCs are not provided.