POET Creator

The system includes convolutional neural network to detect people in real-time from a live camera or video feed. The tracking data can be converted into visual content. The system is fully GDPR compliant.

The system integrates optional professional grade Ai 3D sensors suitable for permanent installations inside and outside buildings. Each sensor provides multiple input parameters such as people counting with optional modules for gender and adult/child. The sensors can be used independently or combined to create a building wide detection area. Each parameter can be used separately as trigger for content engines parameters, content effects and smart scheduling.

POET supports Papago CO2 sensors as input for air quality and any microphone as input for loudness of space.

Network RTPS cameras or USB cameras can be connected and used as live inputs with responsive content effects such as optical flow particles.

The system ships with live data input parameters for weather, calendar and time in any location, as well as powerful selection of data simulators. TCP/UDP and MQTT data streams can be implemented easily. Project specific data harvesting can be implemented as custom integration.

POET can be centralised as the control point for other systems or a part of a third party centralised system by sending & receiving IP based commands from the network. Any configurable IP based TCP/UDP and OSC protocol device can be linked to the system as input or receiver to impact content behaviour or change states in the Scheduler.

Built-in suite of sophisticated visual content engines create high quality real-time generative graphics. These are suited for low resolution pixel surfaces and lighting features as well as high definition streams for screens and projectors. Each content engine has parameters such as speed, colour, direction that can be linked to any data input. The content is rendered in real-time & changes are instantly visible on the output. All of the engines are freely customisable to suit the project.

A full effects menu is available and usable with any type of content. The effects include filters such as: Blurs, Chroma, Hue & Saturation, Color overlays, Content Slicing, Ripple, Rotate, Pixelate, Sharpen, LUT, invert, isolate, mirror, optical flow and more. Each of the effects parameters can be linked independently to external inputs.

All the real-time content engines, video inputs, camera inputs, image content and content effects are complete with rich variety of parameters that can be linked independently to tracking inputs, sensor inputs and data inputs making any parameter of any of the contents real-time responsive.

System can import 3D models with pre-defined UV map of the project and map the visual content accurately onto it displaying the active content in real-time. 3D preview enables snapshots and video recording while moving around the 3D environment with content running.

The system supports pre-rendered video content files with screen size up to 8K and allows them to be modified and triggered real-time with all the standard content effect parameters and live inputs.

Still images with screen size of up to 8K to be imported and used as content with all the standard content effects. It is possible for each image to be modified and triggered real-time with all the standard content effect parameters.

POET allows for pixel precise layout creation for different contents within the full output resolution canvas area. Each canvas can be divided into control groups to work with different outputs or content types simultaneously. All the contents can be freely adjusted and multiple content windows can overlap or layer on top of each other or play on individual zones without any restrictions.

Unlimited number of content presets are easy to create where multiple content can be sent to multiple zones within single preset. Presets have layering function allowing unlimited number of different contents to be run on top each other with alpha channel transparency.

Conditional scheduling is possible where external inputs act as triggers that can be combined with other events. This launches content depending on the various live inputs such as weather conditions, time, date, tracking inputs or sensor data or any combination of the same. Mixing different input triggers is unlimited and available separately for different control groups. The user is able to create complex "behavioral rules" without any scripting or programming skills.

The POET system can be integrated with separate special virtual reality digital twin software. The digital twin displays the POET content correctly in real-time and allows for free navigation as well as easy creation of fly through videos and instant renderings.