Zobrazeno 1 - 5
of 5
pro vyhledávání: '"Sovila Srun"'
Publikováno v:
IEEE Access, Vol 11, Pp 128044-128060 (2023)
Many existing text recognition methods rely on the structure of Latin characters and words. Such methods may not be able to deal with non-Latin scripts that have highly complex features, such as character stacking, diacritics, ligatures, non-uniform
Externí odkaz:
https://doaj.org/article/ed9b51c91500439c9b57783f8d92cef4
Publikováno v:
E3S Web of Conferences, Vol 486, p 03014 (2024)
By the empower of the technology and the Internet, there are huge amount of electronic text documents becoming available from day to day. Seeking information in the field of agriculture numerous collection is required well organized documentations th
Externí odkaz:
https://doaj.org/article/e9e951da59554aaaa0975fd4c7d30191
Publikováno v:
Journal of Imaging, Vol 9, Iss 12, p 276 (2023)
Attention-based encoder–decoder scene text recognition (STR) architectures have been proven effective in recognizing text in the real world, thanks to their ability to learn an internal language model. Nevertheless, the cross-attention operation th
Externí odkaz:
https://doaj.org/article/48db3ac4ce6944a78f43c0ce80ab2d73
Publikováno v:
Journal of Imaging, Vol 9, Iss 11, p 248 (2023)
Connectionist temporal classification (CTC) is a favored decoder in scene text recognition (STR) for its simplicity and efficiency. However, most CTC-based methods utilize one-dimensional (1D) vector sequences, usually derived from a recurrent neural
Externí odkaz:
https://doaj.org/article/111390db972d4728b9e22d6c2daa6231
