7 citations · 20 across the 7 of their papers we have counts for
8 papers
Foveated Compression: Selective High-Resolution Preservation for Token-Efficient VLMs
Donghyun Han, Jangho Park, Yuseok Bae
Visual tokens are a major source of inference cost in vision-language models, yet simple image downsampling remains a surprisingly strong compression baseline. This raises a comple…
PersonaDrive: Controllable Trajectory Prediction with Multi-Dimensional Driving Personas
Chan Lee, Kimin Yun, Yuseok Bae +2
Although recent trajectory prediction and end-to-end autonomous driving methods improve robustness in urban environments, they still lack meaningful controllability. Existing bench…
Learning to Orchestrate Vision Foundation Models for Multi-Task Dense Prediction
Donghyun Han, Yuseok Bae, Jung Uk Kim +1
Vision foundation models (VFMs) exhibit complementary strengths shaped by their pretraining objectives. Yet prevailing methods for multi-task dense prediction still train an entire…
Task-Specific Adaptation of Segmentation Foundation Model via Prompt Learning
Hyung-Il Kim, Kimin Yun, Jun-Seok Yun +1
Recently, foundation models trained on massive datasets to adapt to a wide range of tasks have attracted considerable attention and are actively being explored within the computer…
Adversarial Training with Stochastic Weight Average
Joong-Won Hwang, Youngwan Lee, Sungchan Oh +1
Adversarial training deep neural networks often experience serious overfitting problem. Recently, it is explained that the overfitting happens because the sample complexity of trai…
Localization Uncertainty Estimation for Anchor-Free Object Detection
Youngwan Lee, Joong-won Hwang, Hyung-Il Kim +4
Since many safety-critical systems, such as surgical robots and autonomous driving cars operate in unstable environments with sensor noise and incomplete data, it is desirable for…