Alphanet is an Open Source (To be released soon), the marketplace for pre-trained models, optimized for running on Android Devices
Alphanet is an Open Source (To be released soon), the marketplace for pre-trained models, optimized for running on Android Devices. It uses Tensorflow Mobile for this task. We provide a collection of trained neural network models which you can download and run within the application.Alphanet brings the power of Machine Learning for your Android Device. It has to offer a lot, even though it is an initial release. Alphanet offers the following features.# 1: Ability to take a model offline and run
it without internet connectivity. You can be assured no data leaves your device, it is on device prediction. # 2: A wide variety of Runner UI enables you to run different models from Convolutional to Simple Feed forward models. Currently, we offer two Runner UI's named Raw Input Runner UI and Slate Runner UI. In the future, we will add support for many other Runner UI's# 3: Testing Your models made easy, with Alphanet you can build your own models and train->freeze->optimize and run the model, directly from within the application, Using One of its Runner UI. For More Info Read the FAQ section of the Application# 4: Publish your own models to the world, Soon We will enable the support for Public to release and publish any ML model they wish to the worlds.Following are the list of models we currently offer in the application : 1. Iris Classifier: The Popular Iris Classifier Model, ready to classify is available2. MNIST: Trained on MNIST, We have a Model that is ready to recognize the handwritten digits in Slate View3. EMNIST: Extended MNIST pushes the limits of MNIST and even recognizes all the letters and alphabets including Lower and Upper Case letters.**More will be added in a day or so****WE WILL ADD MORE AND MORE FEATURES TO THE APPLICATION, STAY TUNED**
... moreless ...