Qinnan Shangguan
Papers
1
Total Citations
10
H-Index
1
About
Qinnan Shangguan is a rising researcher in computer vision, with a focus on few-shot object counting and foundation model adaptation. Their most cited work, "Can SAM Count Anything? An Empirical Study on SAM Counting" (2023, 10 citations), investigates the application of Meta AI's Segment Anything Model (SAM) to the challenging task of counting objects from unseen categories with minimal examples. This study systematically evaluates SAM's zero-shot and few-shot counting capabilities, revealing both its strengths and limitations in class-agnostic segmentation for density estimation. Shangguan's contributions provide critical insights into how large-scale vision models can be repurposed for fine-grained numerical tasks, bridging the gap between segmentation and counting. By empirically testing SAM's performance across diverse datasets, their work helps guide future research on adapting foundation models for specialized vision problems. Though early in their career, Shangguan's research has already informed the community's understanding of SAM's versatility, making it a valuable reference for students and researchers exploring few-shot learning and object counting.
Research Focus
Key Achievements
Top Papers
- 1Can SAM Count Anything? An Empirical Study on SAM Counting10 citations · 2023