Cognitive and AI Services for Image Recognition to the Palm of your Hand

28 Jul 2018

Announcing BlueforceEDGE Image Recognition Plugin for superior decision support in time stressed and contested environments

The increasing speed of unpredictable and chaotic events collapses the tactical decision space for military commanders and first responders. Forward sensor fusion with edge-based processing accelerates recognitional decision making by widening the decision-maker’s aperture to a greater number of sensor inputs and a wider array of sensor types. Cognitive computing tools compare signals to patterns to automate decision cues and open the door to meta-cognition, discovery of new patterns, and continuous improvement of decision models.

The new BlueforceEDGE Image Recognition Plugin enhance’s real-time recognition and speed of decision for operators and fixed and mobile command nodes by recognizing persons of interest as well as object recognition.  BlueforceEDGE V3 (“EDGE”) software delivers an extensible autonomous agent software platform that enables sensors from multiple manufacturers to be fused into a secure and interoperable stream of data enabling a single pane of glass view of your entire mission space. EDGE not only extends access to sensors from disparate manufacturers, but also allows development of software plugins providing access to enterprise services, iamge recognition systems, video and data analytics, rules engines, and more.

Blueforce’s new Image Recognition Plugin for BlueforceEDGE extends cloud intelligence to the mobile edge by exposing cognitive services for electro optical imagers (from your mobile device, body worn camera, unattended ground sensor, and/or persistent surveillance systems). BlueforceTACTICAL and BlueforceCOMMAND endpoints may promote imagery (manual or autonomous) to a BlueforceEDGE endpoint that has enabled the Image Recognition plugin. The plugin then pushes the image to the cloud for analysis, monitors the probability scoring, and then routes the result set in seconds to those near the event, or those with a specific need to know.

Version 1 of the BlueforceEDGE Image Recognition plugin exposes the following services:

  • Custom Vision Service: Allows the training and construction of custom image classifiers for object identification such as firearms, knives, drones/UAV, and more.
  • Face Detection: Detect one or more human faces in an image and get back face rectangles and unique IDs for where in the image the faces are, along with face attributes which contain machine learning-based predictions of facial features. Face attribute features include: Age, gender, pose, smile, and facial hair along with 27 landmarks for each face in the image.
  • Face Identification: The system is built with privacy in mind and only detects persons of interest by matching faces in your private repository.
  • Emotion Recognition: Returns a confidence level across a set of emotions for each face in the image such as anger, contempt, disgust, fear, happiness, neutral, sadness, and surprise. These emotions are understood to be cross-culturally and universally communicated with particular facial expressions.
  • Optical Character Recognition: Detect text in an image using optical character recognition (OCR) and extract the recognized words into a machine-readable character stream. Analyzes images to detect embedded text, generate character streams, and enable searching.

BlueforceEDGE has long been leveraged as a autonomous sensor fusion platform, processing unattended ground sensor data and making it shareable in far forward operational environments. It is the component in the Blueforce product arsenal that has truly enabled hybrid cloud and edge computing since 2011. The ability to “containerize” algorithms and push them far forward for edge processing changes the response and sense-making equation across a wide variety of tactical missions that include:

  • Site Exploitation: The main intent of site exploitation is to extract as much potential intelligence as possible from a location of interest in hopes that the information collected mitigates potential threats, but can also answer priority intelligence requirements. To date, much of the collected information must be transmitted to other locations for “recognition” and potential sense-making, leading to time gaps and potential missed opportunities. Cognitive services using edge processing provides recognitional augmentation in near real-time while still at the site.
  • ISR: US Department of Defense officials have estimated that a single sensor in a single combat theatre captures more than three NFL seasons worth of imagery in a single day. Exploitation of this avalanche of data is daunting and the perishability of this data can be measured in hours. Autonomous AI services such as imagery recognition can help pick the needle out of the haystack in far forward environments where the tactical value of the data is immeasurable.

The combination of networked personnel, sensors, and cognitive support services that enhance meta-recognition in time-stressed decision environments can quickly drive value for the commander. Blueforce products enable effective processing of sensor data and raw intelligence incorporating intelligent sensing and automation of multi-sensor data fusion to enhance situational awareness. The BlueforceEDGE Image Recognition Plugin fuses the tactical edge with the intelligent cloud enabling distributed cognition amongst people, sensors, and services and making the promise of the “intelligent edge” and hybrid cloud a reality. For more information or to arrange a demonstration call us at 866-960-0204 or send an email to info@blueforcedev.com.