Abstract | ||
---|---|---|
•An end-to-end system for detection as well as language identification of texts in natural images.•Use of GAN based architecture for precise text region prediction and CNN based architecture for language identification.•Use of a doubling strategy to consider the context, which highly increases the performance of the system.•Performance evaluation of the proposed method on standard as well as in-house scene text datasets. |
Year | DOI | Venue |
---|---|---|
2020 | 10.1016/j.patrec.2020.06.024 | Pattern Recognition Letters |
Keywords | DocType | Volume |
Multi-lingual,Scene Text,MSER,SWT,Deep learning,GAN,CNN | Journal | 138 |
ISSN | Citations | PageRank |
0167-8655 | 3 | 0.39 |
References | Authors | |
0 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Shaswata Saha | 1 | 3 | 0.39 |
Neelotpal Chakraborty | 2 | 6 | 2.12 |
Soumyadeep Kundu | 3 | 9 | 3.22 |
Sayantan Paul | 4 | 9 | 2.21 |
Ayatullah Faruk Mollah | 5 | 33 | 8.59 |
Subhadip Basu | 6 | 385 | 43.75 |
Ram Sarkar | 7 | 420 | 68.85 |