Open neptune4year opened 4 years ago
@neptune4year at training, my guess is that there is a script that creates char_dict.txt
from the dataset annotation.
Close this issue.
@neptune4year From the paper, it seems this model is not able to recognize multi-language characters yet (as you said there are only 68 classes). For 2017 MLT dataset they only have detection task, no recognition task
the paper said that there only 68 classes for character recognition, but for 2017MLT dataset, how do you recognize characters out of English? Thank you.