Real-time face recognition and processing

Cross-platform C++ library that includes all algorithmic functionality for analyzing human face images. Provides maximum efficiency and flexible integration into third-party software. Well suited for integration into video surveillance systems (Video Management System), access control and management systems (ACMS), self-service terminals, mobile applications.

Capabilities

• ●
  Face detection in images
• ●
  Face tracking in video, best frame selection
• ●
  Face image quality assessment (overall, rotation angles, blur, occlusions)
• ●
  Biometric template creation and comparison: verification (1:1 comparison), identification (1:N comparison)
• ●
  Demographic classification by face (gender, age)
• ●
  Liveness detection
• ●
  Face mask detection
• ●
  Attribute classification: presence and type of headwear, glasses, mustache, beard, etc.
• ●
  Additional attribute classification: presence and type of headwear, glasses, mustache, beard, etc.
• ●
  Gaze direction determination
• ●
  Phone conversation detection
• ●
  Smoking detection
• ●
  Deepfake detection

Vehicle recognition in video. Determines type, color, make, license plate

Cross-platform C++ library implementing vehicle image analysis functions. Provides maximum efficiency and flexible integration into third-party software. Well suited for integration into video surveillance systems (Video Management System) and access control and management systems (ACMS).

Capabilities

• ●
  License plate recognition in images and video
• ●
  Number recognition with readability assessment and classification by type and country
• ●
  Vehicle silhouette detection and tracking in video
• ●
  Vehicle category determination
• ●
  Special vehicle detection
• ●
  Vehicle make, model and color determination
• ●
  Driver seatbelt detection
• ●
  Mobile phone usage detection while driving

Human detection and tracking in video, attribute classification, re-identification

Cross-platform C++ library designed for human detection in images. Also includes web service for web API integration.

Capabilities

• ●
  Human silhouette detection and tracking in video stream
• ●
  Re-identification – person search by sample based on shape, silhouette, clothing, etc.
• ●
  Multi-camera tracking
• ●
  Heat map generation
• ●
  Attribute classification: clothing, headwear, bags, etc.
• ●
  Fall detection
• ●
  Smoking detection

Recognition of passports, driver's licenses and other standard documents

Cross-platform C++ library implementing standard document recognition functions. Provides maximum efficiency and flexible integration into third-party software. Well suited for integration into access control and management systems (ACMS), self-service terminals. Also includes web service for web API integration.

Capabilities

• ●
  Russian Federation passport
• ●
  CIS ID card
• ●
  Russian SNILS
• ●
  Russian driver's license
• ●
  Vehicle registration certificate (VRC)

Assessment and analysis of people presence in frame

Cross-platform C++ library and web service based on it for estimating number of people in frames. Allows counting people even at very low resolution (from several pixels).

Capabilities

• ●
  Venue attendance monitoring
• ●
  Room occupancy fluctuation analytics
• ●
  Public place overcrowding control
• ●
  People interest zone detection

Real-time operator state analytics and attention monitoring

Web service and C++ library for real-time monitoring of operator and employee state. Provides tools for detecting signs of fatigue, distraction, dangerous behavior and misconduct. Ensures maximum efficient integration into third-party software: video management systems (VMS), workforce management systems (WFM), HR systems and analytics platforms. Suitable for integration into contact centers, dispatch centers, security complexes, vehicle fleets and production facilities.

Capabilities

• ●
  Face detection in frame and tracking in video stream
• ●
  Fatigue detection: recognition of yawning, closed eyes, characteristic head movements
• ●
  Attention distraction detection: analysis of gaze direction and head trajectory
• ●
  Dangerous behavior classification: smoking, eating, drinking, mobile phone usage
• ●
  Personal protective equipment detection (helmet, vest, glasses)
• ●
  Statistics and reporting generation by time and violation types
• ●
  Integration with alerting systems for operational response

Automatic real-time recognition and tracking of railway rolling stock

The intelligent RAILWAY SDK solution is designed for the identification and monitoring of railcars, locomotives, and railway platforms. The system is powered by advanced computer vision algorithms, enabling highly accurate recognition even under challenging operating conditions such as low lighting, extreme weather, or heavily worn markings.

Capabilities

• ●
  Multi-format recognition: automatic reading of 8-digit freight wagon numbers (including check-digit validation), two-line passenger carriage numbers, and non-standard locomotive numbering formats
• ●
  Operation in motion: reliable recognition at train speeds from 2 to 50 km/h, including correct handling of reverse movement, stops, and direction changes during shunting operations
• ●
  Robustness to external factors: effective performance at dusk and in adverse weather conditions. The algorithms compensate for barrel distortion caused by wide-angle lenses and reliably read dirty or partially damaged markings
• ●
  Intelligent aggregation: object tracking within the video stream and fusion of results from multiple cameras (left/right side views and undercarriage view) to ensure maximum data reliability
• ●
  Optical identification: the solution operates entirely on video analytics and does not require the installation of external wagon start/end sensors or magnetic loops
• ●
  High accuracy: the recognition success rate exceeds 92% when using a single camera and reaches 96–98% with dual-side capture (figures are approximate and based on internal estimates)