Qianqian Shen
Papers
1
Total Citations
2
H-Index
1
About
Qianqian Shen is a rising researcher in computer vision, with a primary focus on advancing instance detection and open-world visual understanding. Her most-cited work, "Solving Instance Detection from an Open-World Perspective" (2025, 2 citations), tackles the challenging task of instance detection (InsDet)—localizing specific object instances in novel scenes using visual references. She identifies a critical gap: existing methods assume closed-world scenarios where all query instances are present, failing in open-world settings where many objects are unseen. Her major contribution is a novel framework that decouples instance detection into two robust stages—open-world proposal detection to identify all potential objects, followed by instance-level matching to pinpoint those of interest. This approach significantly improves generalization to unseen instances, a key step toward real-world deployment. While her citation count is early-stage, her work addresses a fundamental limitation in object recognition, positioning her as an innovator in open-world vision systems. Her research promises to enhance applications from robotics to surveillance, where encountering novel objects is the norm.
Research Focus
Key Achievements
Top Papers
- 1Solving Instance Detection from an Open-World Perspective2 citations · 2025