GV-AI Server is a video analytic software designed to provide various VA features at a retail store for up to 8 channels of IP cameras. It is able to identify and compile customer statistics, such as age, gender and customer behavior for sales maximization, as well as automatically notify managers or security personnel of inventory shortages or upon detecting suspicious persons.
Features:
- Support for up to 8 channels of IP cameras
- Realtime face recognition and tracking
- Realtime masked face recognition and tracking
- Average recognition speed of within 1 second per face when the recognition targets are moving toward the cameras
- Face profiling by age and gender
- Product Attention with heatmap display for monitoring product interest by foot traffic
- Short Inventory monitoring and alert
- Queue Management for monitoring the cashier checkout time and the number of persons waiting in line
- Suspect & Loitering Detection
- Human Counter for counting the number of people that enter and exit at the vicinity
- Support for AI-capable GV-IP cameras and GV-3D People Counter V2 for collecting people flow data In / Out / Stay
- Automatic compilation of statistical graphs for each type of VA event
- Exportable daily / weekly / monthly / yearly hourly face recognition count
- Query by VA events
- Integration of GV-VMS for video recording and playing back face recognition events
- Master and Slave feature for interconnecting multiple GV-AI Servers with a centralized Face Database
SPECIFICATIONS
|
Video Input |
Up to 8 channels |
|
|
Video Codec |
H.264, H.265 |
|
|
Video Resolution |
Up to 12 Megapixels per channel |
|
|
IP Camera Supported |
GV-Face Detection Cameras (GV-BX2600-FD / 2700 / 2700-FD / 4700 / 4700-FD / 8700-FD, GV-MD8710-FD), GV-Fisheye Cameras, IP cameras via ONVIF or RTSP, USB webcams |
|
|
Networking |
LAN, WAN, Internet |
|
|
Video Analysis |
Functions |
Face Attributes, Face Recognition, Product Attention, Short Inventory Alert, Queue Management, Suspect & Loitering Detection, Human Counter |
|
Face Recognition Speed |
Within 1 second per face when the recognition targets are moving toward the cameras |
|
|
No. of Face Profiles |
100,000 |
|
|
No. of Face Groups |
Unlimited |
|
|
No. of GV-AI Server Interconnectable |
11 (1 Master + 10 Slaves, sharing 1 Face Database managed by the Master) |
|
|
VA Event Trigger |
I/O Box Output, LINE Notification, Telegram Notification, Trigger Camera Output, Push Notifications from Mobile App, E-mail Notification |
|
|
Software Integration |
GV-VMS V18.1 or later (video surveillance) |
|
|
Mobile App |
GV-Assistant app GV-Welcome for Android TV boxes / mobile devices (for Welcome screen display) |
|
|
Language |
English, Traditional Chinese |
|
