PROJECT TITLE :
Strokelets: A Learned Multi-Scale Mid-Level Representation for Scene Text Recognition
In this paper, we have a tendency to are involved with the matter of automatic scene text recognition, which involves localizing and reading characters in natural pictures. We tend to investigate this downside from the perspective of representation and propose a completely unique multi-scale representation, which results in accurate, sturdy character identification and recognition. This representation consists of a collection of mid-level primitives, termed strokelets, which capture the underlying substructures of characters at different granularities. The Strokelets possess four distinctive benefits: one) usability: automatically learned from character level annotations; 2) robustness: insensitive to interference factors; 3) generality: applicable to variant languages; and 4) expressivity: effective at describing characters. Intensive experiments on commonplace benchmarks verify the advantages of the strokelets and demonstrate the effectiveness of the text recognition algorithm designed upon the strokelets. Moreover, we tend to show the method to include the strokelets to improve the performance of scene text detection.
Did you like this research project?
To get this research project Guidelines, Training and Code... Click Here