OpenPTrack V2 “Gnocchi” coming in August!


The OpenPTrack team is excited to announce the upcoming OPT V2 (Gnocchi) release—planned for mid-August—which provides new, fundamental features:

  1. Object tracking. For the first time, OpenPTrack will track objects in addition to the human body. V2 will add the capability to track objects based on color profile, a feature we expect to extend to other types of detectors in 2017-2018.
  1. Pose recognition. OPT V2 will now add the capability to annotate person tracks with poses recognized from a pre-trained set, using machine learning-based skeletal tracking. (GPU acceleration required.)
  1. Enhanced real-time person detection. OPT V2 includes a new, experimental person detector using convolutional neural networks; this improves performance in detecting people who are sitting, lying down, heavily occluded, etc. (GPU acceleration required.)
  1. ZED camera support. OPT V2 integrates support for the Stereolabs ZED camera, enabling the system to operate in high ambient light conditions where the Kinect and similar imagers fare poorly. OpenPTrack is going outside!

The Gnocchi release also updates the underlying software stack to Ubuntu Linux 16.04 LTS and Robot Operating System (ROS) Kinetic Kame. (Though some components will still work with Ubuntu 14.04 and ROS Indigo, we will no longer be able to test/support them.)

OPT V2 Gnocchi has been under development for the past year. Our first planned deployment will be at UCLA Lab School for the ongoing NSF-supported iSTEP project.

 

(Click any image for larger views.)

Multiview Pose Tracking

Multiview Pose Tracking

Object

Object Detection

Zed Camera, CNN-based People Detection

Zed Camera, CNN-based People Detection