search

tantara / JejuNet

Real-Time Video Segmentation on Mobile Devices with DeepLab V3+, MobileNet V2. Worked on the project in 🏝 Jeju island
http://jeju.dlcamp.org/2018/
MIT License
279 stars 54 forks source link

How is model trained and converted? #3

Closed haikuoyao closed 5 years ago

haikuoyao commented 5 years ago

How is Deeplab model trained and converted to TF-Lite file?

Thanks.

RizanPSTU commented 5 years ago

Please provide the answer. I am requesting

tantara commented 5 years ago

@haikuoyao @RizanPSTU Hi guys, please refer to the official tutorials https://github.com/tensorflow/models/blob/master/research/deeplab/g3doc/quantize.md