Computer Vision in Security, Surveillance, Private and Public Safety
System automatically detects threat levels originating from hundreds of video streams (CCTV cameras, dash cam / drone footage etc). Algorithm weighs threat levels according to object & scene classification, quantity, velocity and other factors. User interface automatically alerts human operator and zooms into streams requiring immediate attention. System increases situational awareness, reduces reaction time, reduces human labour and hardware requirements in both real-time surveillance and historical footage analysis. Read more..
System detects and stores number plates in vicinity of your property or your vehicles. Robo advisor notifies you of any anomalies that are worth attention such as frequent appearance or suspicious behaviour of certain non-whitelisted vehicles. Works with dash cameras and CCTV cameras.
Aerial UAV, satellite, bird eye view analysis using specialized computer vision models, embedded deep learning hardware and DeepX triage technology.
System consists of deep learning models trained to recognise different types of cooked food and energy conversion database allowing to effectively estimate calories intake and dietary pattern using photo and video recordings of consumed meals. Read more..
Our solution allows to detect objects within aerial UAV / satellite imaging (both static and dynamic) faster, cheaper and more efficiently in comparison to human eye detection.
First responders auto alert system (computer vision powered)
Using unique DeepX technology it is possible to detect events posting threats to health and lives of the public (and/or infrastructure damaging) immediately as they occur, alerting first responders immediately. This saves valuable seconds and minutes in reaction time resulting potentially in lives saved and reduction of damages to public health and infrastructure. Examples of events could be fires, explosions, natural disasters, car crashes, industrial accidents, criminal and terrorist activity. DeepX team has developed efficient machine learning models for these types of events.
DeepFake is on the rise lately in the entertainment sector, however it poses a serious security threat when applied by criminals. It can also be used to swing public opinions and moods to support fake news production. DeepX team is experienced in deep fake technology and explains how machine learning is used to detect and protect against deep fake products where human brain may fail.
Computer vision for head pose and line of sight estimation
Line of sight, head pose / direction and eye tracking are important input factors in multiple use cases starting from security to gaming, sports, entertainment, 3D / VR / augmented reality, retail, commerce, advertising, healthcare and so on. With primary focus on security DeepX offers universal capability allowing to apply head pose, eye tracking, line of sight estimation sometimes combined with other types of detections (such as face recognition and gait recognition) to solve real-world problems.
Computer Vision powered Face recognition
Face recognition. DeepX applies state of the art technologies including machine learning software and hardware to provide best in class real-time face detection capability to be used in security, public safety, commerce, workspace analytics and other purposes.
Machine speed deflection and projection (computer vision powered)
Computer vision and robotics can be used to warn and defend individuals, public and equipment against physical attacks. Robotic deflection, counter-attack and distraction / vision blockage (by way of lasers, for example) can be used efficiently when paired with real-time machine speed computer vision detection capability. DeepX leads the research in this area.
VIP AI / Computer Vision security system (AI Security team)
As an evolution of Deep Alerts Triage project, DeepX offers a package for VIP security that could be used to boost personal or business security. In real world applications huge amounts of CCTC surveillance and other camera and sensor data is lost without extracting all useful knowledge from them. This happens due to lack of time, human resources, distractions caused by false positive detection etc. DeepX platform addresses these issues ensuring that “AI Security Officer” takes into account every second of video and imagery data analyzing and alerting the owners in case of any suspicious or threatening activity. The system integrates with existing security and surveillance hardware and software systems already in place, increasing the overall security and situational awareness without creating any further distractions.
Computer Vision in Retail and Commerce
Computer vision for people counting and retail analytics
In retail and event management applications it is often important to count and analyze visitors, attendees and customers behaviour. DeepX offers these capabilities enabling new level of insights and real-time added value generation for example via focused guidance, adaptive merchandising and advertising.
Computer Vision in Robotic navigation and Warehouse management
Computer vision assisted indoor navigation, coordinates and distance estimation
In the absence of reliable GPS coverage and to achieve higher accuracy in indoor applications, automated systems rely on Computer Vision and lidar powered systems in indoor navigations, room dimensions estimation, obstacles detection, coordinates and speed evaluation. DeepX offers computer vision and supplementary technologies powered capabilities to address these problems.
DeepX Professional Services team offers configuration and consulting around Elastic Search stack, in particular Kibana visualizations and Kibana Machine Learning modules, including configuring the data ingestion, indexes, jobs and tools such as Kibana ML Anomaly Detection, Forecasting etc.
Computer Vision in Space exploration
Rank & Sync AI image processing system for satellite imagery
System applies computer vision and AI to facilitate on-board decision making in processing, storage and transmission of satellite imagery. This helps to increase efficiency when working with real-world imaging, storage and bandwidth limitations.