Sagar Sharma short introduction to TensorFlow on mobile. TensorFlow Lite is a lightweight and a next step from TensorFlow Mobile. You can do almost all the things that you do on TensorFlow mobile but much faster.
Just like TensorFlow Mobile it is focused on the mobile and embedded device developers, so that they can make next level applications on systems like Android, iOS, Raspberry PI etc.
What is included with TensorFlow Lite:
- Android neural networks API and C++ API
- New model file format .tflite
- Speed: new neural network API makes computation much faster
- Privacy: the data does not leave the device
- Availability: it works offline
- No computation cost: computation on your device, no server needed
- Pre-tested models: models work out of the box
There are some caveats as well. You will need to pay attention to the size of your models. Models may take up multiple megabytes of space. Evaluating neural networks involve a lot of computation, which could increase battery power usage.
Links to further reading, resources and explanation charts also included in this article. Nice reading!
[Read More]