project: OCR author: cwx
视频字幕识别
tensorflow1.6
pytorch0.4
CUDA9.0
CUDNN7.0
python3.5 或者 3.6
安装lmdb: pip install lmdb
(a) cd ctpnsource/lib/utils and execute: python setup.py build
(b) copy the .so file from the "build/XXX" directory to the ctpnsource/lib/utils
sh setup.sh
然后输入文件名,如:
ren_1.jpg