Visual Understanding in the Open World

Open world vision refers to the ability of computer vision systems to perceive, understand, and adapt to dynamic and unpredictable real-world environments. Unlike traditional closed-set models that operate within fixed categories and controlled settings, open world vision emphasizes generalization to unseen objects, continual learning, and robust decision-making in the presence of uncertainty. This paradigm is critical for enabling AI to function reliably in complex applications such as autonomous driving, robotics, and augmented reality.

Publications
1. Wentao Bao, Kai Li, Deep Anil Patel, Yuxiao Chen, and Yu Kong. Exploiting VLM Localizability and Semantics for Open Vocabulary Action Detection. Winter Conference on Applications of Computer Vision (WACV), 2025. Paper
2. Wentao Bao, Lichang Chen, Heng Huang, and Yu Kong. Prompting Language-Informed Distribution for Compositional Zero-Shot Learning. European Conference on Computer Vision (ECCV), 2024. Paper
3. Xiwen Dengxiong and Yu Kong. Ancestor Search: Generalized Open Set Recognition via Hyperbolic Side Information Learning. IEEE/CVF Winter Conferences on Applications of Computer Vision (WACV), 2023.
4. Wentao Bao, Qi Yu, and Yu Kong. OpenTAL: Towards Open Set Temporal Action Localization. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022. Oral. Paper, Code
5. Wentao Bao, Qi Yu, and Yu Kong. Evidential Deep Learning for Open Set Action Recognition. International Conference on Computer Vision (ICCV), 2021. Oral. [Paper, Project]
6. Wentao Bao, Qi Yu, and Yu Kong. DRIVE: Deep Reinforced Accident Anticipation with Visual Explanation. International Conference on Computer Vision (ICCV), 2021. [Paper, Project]