Ajay J. Joshi
Papers
3
Total Citations
43
H-Index
3
About
Ajay J. Joshi is a leading researcher in computer vision and machine learning, with a focus on active learning strategies for image classification and video analysis. His work addresses critical challenges in robotics and surveillance, where accurate object recognition is essential despite limited training data. Joshi’s most-cited paper, "Multi-class batch-mode active learning for image classification" (2010, 22 citations), introduces a framework that intelligently selects the most informative samples for labeling, significantly reducing the data required for training robust classifiers. He further advanced the field with "Learning of moving cast shadows for dynamic environments" (2008, 14 citations), where he pioneered an online statistical learning approach using support vector machines to distinguish moving shadows from foreground objects—a key problem in video surveillance. His research on "Coverage optimized active learning for k-NN classifiers" (2012, 7 citations) extends these ideas to fast, real-time classification systems for robotics applications like exploration and rescue. With a cumulative impact of over 43 citations across his top works, Joshi’s contributions have shaped efficient, data-driven methods that enable autonomous systems to perceive and interact with dynamic environments more effectively.
Research Focus
Key Achievements
Top Papers
- 1Multi-class batch-mode active learning for image classification22 citations · 2010
- 2Learning of moving cast shadows for dynamic environments14 citations · 2008
- 3Coverage optimized active learning for k - NN classifiers7 citations · 2012