* Cognika - Perseus
Description:
Real-Time Alerting and Visual Search for Fixed and PTZ Cameras
Features:
Real-Time alerting on humans or vehicles; Content-based image search on humans or vehicles
Multi GPU support:
Yes
Real-Time Alerting and Visual Search for Fixed and PTZ Cameras
Real-Time alerting on humans or vehicles; Content-based image search on humans or vehicles
Yes
GPU accelerated decode & rendering enables the display of more high-resolution streams from a single workstation, as well as enhancing video playback performance.
Offloads the workload of video stream decode and display rendering of multiple streams to GPU.
Single only
Real time facial recognition and forensic alerts against multiple watchlists.
Supports crowded scenes, difficult lighting, faster than real-time analysis, partial face concealment.
Yes
Intelligent analysis of video on 1,000+ camera streams to significantly filter and reduce the camera streams requiring an operator view.
GPU accelerated machine learning to identify abnormal activity within video streams
Yes
Real-time alerts and data reporting from use cases include Security, Traffic, Retail and Parking, Analysis of video streams in real-time and on archived video at up to 20x real-time speeds.
Robust and user trainable object classification for tracking. Using distributed architecture
Yes
Deep Learning facial recognition SDK with 25 bytes template for real-time identification applications and large-scale IdM solutions
Real-time face detection, verification or suspect identification against multi-million datasets based on an artificial neural network
Yes
Open-platform video management software for scalable IP video surveillance systems with advanced video analytics.
H264 decoding for CPU offload, zooming, image conversion shader from 24 to 32 bits delivering better color combination
Yes
Accelerated library for video analysis on video surveillance.
Accelerated Intrusion Detection Algorithm.
Yes
Real-time (render-less) super-resolution- based video enhancement and redaction software for forensic analysts and law enforcement professionals
Multi-filter, render-less video reconstruction (super-resolution, stabilization, light/color correction), and automatic tracking for redaction video from body cameras, CCTV and other sources.
Yes
Face recognition for real-time video surveillance and offline search compared against multiple watch-lists.
Detects & recognizes multiple faces simultaneously in crowds and variable lighting, scales to more cameras, larger face databases.
Yes
High speed visual search and analysis
Uses images instead of keywords to search for objects or scenes of interest within video and imagery. Reliant solely on pixel data with no training, keywords, or tags required.
Yes
IP video management system designed for auto discovering, managing, recording, analyzing and searching thousands of video streams at the same time
GPU accelerated conversion of YUV images to RGB, drawing and scaling YUV images in desktop client, dewarping fisheye (circular or panamorph) live or recorded video streams
Yes
Real-time face recognition in cooperative and uncooperative scenarios adaptable for a multitude of applications to detect, identify or verify people and objects.
Critical core segments written in CUDA allowing for unlimited parallelization and transparent clustering.
Yes