TF Multi Classifier

edodm85

Contains adsIn-app purchases

About this app

TF Multi Classifier App uses Tensorflow lite to classify camera frames in real time.
The results are displayed on the top of the screen.

The App supports these processing hardware:
- CPU (all phones)
- NNAPI (requires Android 8.1 or higher)
- GPU (requires OpenGL ES 3.1 or higher)

The App starts with a default google dataset but if you want you can insert a custom dataset.

The defaul dataset is: quantized Mobilenet model (mobilenet_quant_v1_224).

You can download the following datasets from here: http://bit.ly/2BvvDNT
- Dog Breed (created with Transfer learning)
- Dog Vs Cat (created with Transfer learning)
- Bird (converted from google pb dataset)
- Insect (converted from google pb dataset)
- Plant (converted from google pb dataset)
- Seefood (converted from google pb dataset)


You can find my tutorial here: http://bit.ly/2V3giK9




NB: For custom App contact me.

Please rate and review so I can make it better!

What’s new

v2.5.5
- Various improvements

v2.5.4
- Updated launcher icon
- Added NPU support
- Fixed some bugs

v2.5.2
- Updated LiteRT to v2.1.6
- Added new training model
- Removed unsupported NNAPI delegate
- Refreshed GUI graphics
- Updated Android target SDK to API 37