ypwhs/baiduyun_deeplearning_competition
A deep learning project using recurrent neural networks with CTC Loss to crack captchas and recognize arithmetic expressions.
Not currently ranked — collecting fresh signals.
star history
This repository contains Jupyter notebooks for a deep learning competition focused on sequence recognition tasks. It implements captcha cracking using convolutional and recurrent neural networks with CTC (Connectionist Temporal Classification) Loss, which enables training without needing alignment between input and output sequences. The project also includes solutions for recognizing arithmetic expressions at both preliminary and final competition stages, using GRU networks for sequential data processing.
Frequently asked
- What is ypwhs/baiduyun_deeplearning_competition?
- A deep learning project using recurrent neural networks with CTC Loss to crack captchas and recognize arithmetic expressions.
- Is baiduyun_deeplearning_competition open source?
- Yes — ypwhs/baiduyun_deeplearning_competition is open source, released under the MIT license.
- What language is baiduyun_deeplearning_competition written in?
- ypwhs/baiduyun_deeplearning_competition is primarily written in Jupyter Notebook.
- How popular is baiduyun_deeplearning_competition?
- ypwhs/baiduyun_deeplearning_competition has 421 stars on GitHub.
- Where can I find baiduyun_deeplearning_competition?
- ypwhs/baiduyun_deeplearning_competition is on GitHub at https://github.com/ypwhs/baiduyun_deeplearning_competition.