Joydeep Kundu
Papers
1
Total Citations
2
H-Index
1
About
Joydeep Kundu is a researcher specializing in computer vision and natural language processing, with a particular focus on scene text detection and recognition. His most-cited work, "Text Region Identification from Natural Scene Images Using Semi-Supervised MSER Method" (2022), introduces an innovative approach that leverages semi-supervised learning to enhance the Maximally Stable Extremal Regions (MSER) algorithm for accurately localizing text in complex, real-world images. This contribution addresses a critical challenge in autonomous systems, document analysis, and assistive technologies, where robust text extraction from cluttered backgrounds is essential. While his citation count is modest, the work demonstrates a strong methodological foundation in combining unsupervised feature extraction with supervised refinement, offering a scalable solution for low-resource scenarios. Kundu’s research bridges the gap between traditional image processing and modern machine learning, making his findings relevant for students and engineers developing intelligent vision systems. His ongoing efforts promise to advance the reliability of text recognition in dynamic environments, with potential applications in augmented reality and smart city infrastructure.
Research Focus
Key Achievements
Top Papers
- 1