[SUGGESTIONS] G-Colab and better text detection

[MarcoRavich]

Hi there, I didn't tested it yet (since I don't own any discrete GPUs in my main hw configs) but your software seems cool.

It would be great for people like me to have a Google-Colab version of it.

Check out these in order to have an idea:

• https://colab.research.google.com/github/AlphaAtlas/VapourSynthColab/blob/master/VapourSynthColab.ipynb
• https://colab.research.google.com/drive/1arRAJUPBC3a3l2iGghbs3CqDjdD1IQ6y

You can build your own (of course) but you can also ask to add it to MixLab:

• https://colab.research.google.com/github/shirooo39/MiXLab/blob/master/MiXLab.ipynb

Last but not least, it would be great to establish some kind of collaboration with these projects in order to achieve a better text detection:

• https://github.com/YaoFANGUK/video-subtitle-extractor
• https://github.com/shawnsky/extract-subtitles
• https://github.com/Dinghow/OCR_on_Video
• https://github.com/krohak/Embedded-Subtitles-OCR

Hope that inspires !

[zyddnys]

I am working a completely new version, currently only support single image, https://github.com/zyddnys/manga-image-translator
I will use its superior text detection and OCR capability to create a new version that works for videos.

[MarcoRavich]

BREAKING
Seems @oliverfei's videocr-PaddleOCR implemented a Colab version, finally:

https://colab.research.google.com/github/oliverfei/videocr-PaddleOCR/blob/master/videocr_PaddleOCR.ipynb

Hope that inspires !