Products :: TensorFlow :: Overview

Machine Learning. Artificial Intelligence. Perl.

Perl TensorFlow

Perl TensorFlow is a programming interface which allows Perl developers to utilize Google's TensorFlow machine learning platform natively in their own Perl source code.

Features of the Perl TensorFlow interface include:

  • 100% Compatible with Google TensorFlow
  • Load & Run Neural Network Models
  • Classify Unknown Objects
  • Predict Future Values
  • Native Perl API

Products :: TensorFlow :: Code

Pure TensorFlow. Pure Perl.

Perl TensorFlow Code

Perl TensorFlow provides an easy-to-follow Perl TensorFlow Tutorial with real Perl source code examples and a runnable demo of image classification. The tutorial demonstrates a pre-built neural network model, utilizing the MobileNet V2 architecture and trained on the Imagenet dataset.

Features of the Perl TensorFlow Tutorial include:

  • Available as both POD & Jupyter Notebook
  • Real runnable Perl source code
  • Download neural model from Google
  • Download test images from Wikipedia
  • User-friendly graphical output mode

Products :: TensorFlow :: Image Classification

Photos. Automatically Identified.

Perl TensorFlow Image Classification

The Perl TensorFlow Tutorial produces output showing the top 5 potential classifications of each input test image. Where possible, specific plant & animal breeds are automatically determined.

Each potential image class is given a statistical probability ranging from 0 (impossible) to 1 (certain).
All probabilities for a given image should add up to a total value of 1.0, representing 100% certainty overall.

Machine learning image classification can be used for a wide variety of academic & commercial applications, such as machine vision, warehouse inventory control, biomedical research, facial recognition, and wildlife conservation, to name just a few.

Features of the Perl TensorFlow Image Classification includes:

  • Impressive visual demonstration of machine learning
  • Clickable links to the original Wikipedia input files
  • Statistical probabilities for each possible classification
  • Variety of pre-defined sample images
  • Easy to read graphical formatting