Awesome

Open Source TASS AI Projects

tassAI is computer vision software that is able to communicate autonomously with applications and devices via the Internet of Things. There are several versions of TASS and several different projects that evolved from the concept. Each version of TASS uses different techniques and technologies to accomplish facial recognition and other computer vision uses.

This repository provides the history of the open source TASS projects. Each tutorial provides full instructions and source codes.

Latest Facenet Version 0.0.1

The latest version of TASS is TASS Facenet Classifier Version 0.0.1 which uses Siamese Neural Networks and Triplet Loss to classify known and unknown faces. The project uses an UP2, the Intel Movidius and the iotJumpWay for IoT connectivity.

PROJECT LINK: TASS Facenet Classifier Version 0.0.1

Latest Inception V3 Version 0.0.1

The TASS Inception V3 Classifier Version 0.0.1 uses a pretrained or custom trained Inception V3 model and an Intel® Movidius to carry out object and facial classification, both locally and on a live webcam stream. TASS Movidius uses the iotJumpWay for IoT communication and publishes messages to the broker when an object/face is identified.

PROJECT LINK: TASS Inception V3 Classifier Version 0.0.1

TASS Movidius Yolo Classifier

The TASS Movidius Yolo Classifier uses a pretrained Yolo model and an Intel® Movidius to carry out object, both locally and on a live webcam stream. TASS Movidius uses the iotJumpWay for IoT communication and publishes messages to the broker when an object is identified.

PROJECT LINK: TASS Movidius Yolo Classifier

Intel® AI DevCloud TASS Trainer

This was the fourth version of TASS to be open sourced. The AI DevCloud TASS Trainer replicates the transfer learning side of the original program and is trained on the Intel AI DevCloud HPC Cluster (Colfax Cluster). You can use this trainer for many of the computer vision examples you will find in this and related repos.

PROJECT LINK: Intel® AI DevCloud TASS Trainer

Intel® Computer Vision SDK Windows Examples

This was the third version of TASS to be open sourced. The tutorial helps you build a Computer Vision security system on Windows using Intel® Computer Vision SDK, an Intel® Realsense camera/web/ip cam and an Intel® Edison connected to the Internet of Things via TechBubble Technologies iotJumpWay.

PROJECT LINK: Intel® Computer Vision SDK Windows Examples

OpenFace TASS Computer Vision Example

This was the second version of TASS to be open sourced. The tutorial uses the iotJumpWay Python MQTT Library for communication, an Intel® NUC DE3815TYKE or any Linux Desktop running Ubuntu, 1 or more IP Cameras, an optional Realsense camera, and our own deep learning neural network based on the popular OpenFace facial recognition toolkit.

PROJECT LINK: OpenFace TASS Computer Vision Example

Facial Recognition & Identification on Raspberry Pi

This was the first version of TASS to be open sourced and became quite popular on Hackster.io. The tutorial helps you to build a Raspberry Pi that allows you to train a Haarcascades model, detect recognized/unknown people, optionally monitor the camera in near real-time via a stream, and communicate with the iotJumpWay sending sensor and warning messages that will allow your device to autonomously communicate with other IoT devices on your iotJumpWay network.

PROJECT LINK: Facial Recognition & Identification on Raspberry Pi

InceptionFlow IoT Computer Vision Using Inception V3

This is a Python wrapper based around the third version of TASS and Tensorflow. InceptionFlow is an object & facial recognition Python wrapper for the Tensorflow Imagenet example and integrates IoT connectivity using the TechBubble iotJumpWay Python MQTT client.

PROJECT LINK: InceptionFlow IoT Computer Vision Using Inception V3

Bugs/Issues

Please feel free to create issues for bugs and general issues you come across whilst using this, or any other related iotJumpWay issues. You may also use the issues area to ask for general help whilst using the iotJumpWay in your IoT projects.