Výsledky vyhledávání - "Jung, Jiyoung"

Report

Enhancing Visual Classification using Comparative Descriptors

Autor: Lee, Hankyeol, Seo, Gawon, Choi, Wonseok, Jung, Geunyoung, Song, Kyungwoo, Jung, Jiyoung

The performance of vision-language models (VLMs), such as CLIP, in visual classification tasks, has been enhanced by leveraging semantic knowledge from large language models (LLMs), including GPT. Recent studies have shown that in zero-shot classific

Externí odkaz: http://arxiv.org/abs/2411.05357

Zobrazit plný text záznamu

Report

Robust Adaptation of Foundation Models with Black-Box Visual Prompting

Autor: Oh, Changdae, Seo, Gyeongdeok, Jung, Geunyoung, Cheng, Zhi-Qi, Choi, Hosik, Jung, Jiyoung, Song, Kyungwoo

With the surge of large-scale pre-trained models (PTMs), adapting these models to numerous downstream tasks becomes a crucial problem. Consequently, parameter-efficient transfer learning (PETL) of large models has grasped huge attention. While PETL m

Externí odkaz: http://arxiv.org/abs/2407.17491

Zobrazit plný text záznamu

Report

Flat Posterior Does Matter For Bayesian Model Averaging

Autor: Lim, Sungjun, Yeom, Jeyoon, Kim, Sooyon, Byun, Hoyoon, Kang, Jinho, Jung, Yohan, Jung, Jiyoung, Song, Kyungwoo

Bayesian neural network (BNN) approximates the posterior distribution of model parameters and utilizes the posterior for prediction via Bayesian Model Averaging (BMA). The quality of the posterior approximation is critical for achieving accurate and

Externí odkaz: http://arxiv.org/abs/2406.15664

Zobrazit plný text záznamu

Report

Aperiodicity is all you need: Aperiodic monotiles for high-performance composites

Autor: Jung, Jiyoung, Chen, Ailin, Gu, Grace X.

This study introduces a novel approach to composite design by employing aperiodic monotiles, shapes that cover surfaces without translational symmetry. Using a combined computational and experimental approach, we study the fracture behavior of compos

Externí odkaz: http://arxiv.org/abs/2309.05819

Zobrazit plný text záznamu

Report

BlackVIP: Black-Box Visual Prompting for Robust Transfer Learning

Autor: Oh, Changdae, Hwang, Hyeji, Lee, Hee-young, Lim, YongTaek, Jung, Geunyoung, Jung, Jiyoung, Choi, Hosik, Song, Kyungwoo

With the surge of large-scale pre-trained models (PTMs), fine-tuning these models to numerous downstream tasks becomes a crucial problem. Consequently, parameter efficient transfer learning (PETL) of large models has grasped huge attention. While rec

Externí odkaz: http://arxiv.org/abs/2303.14773

Zobrazit plný text záznamu