STAI (Spatial Temporal Artificial Intelligence)

STAI is an artificial mind with awareness of spatial, temporal, and causal relations for processing complex events where every object is anonymously identifiable and trackable at any time in the real world.

STAI is a Smart Video Reasoning and Analytics (SVRA) solution based on a bionic computing architecture that designed to real-time perceive, detect, recall, track, predict, or reason meaningful events associated with living or non-living objects over a vast area with multiple CCTVs and IoT sensors.

STAI supports and automates three functions:
(1) Complex Event Processing
(2) Track and Trace by AI-based Reasoning
(3) Causality and Association Analysis

STAI answers the following six key questions about the objective reality in the physical world:

What happened?
When and Where
Who was/were involved?
Where is/are the people now?
What will happen next?
How to prevent
something bad
from happening?
STAI is now being used in all possible areas.

STAI is used not only by physical security and law enforcement professionals, but by a variety of operators, from marketing to operations, facility management. The ultimate goal is to improve the quality of the living environment in security, safety, living quality and friendly community

  • • Airport or station attack
  • • Entering restricted areas
  • • Criminal behavior, such as shoplifting, burglary, child abduction
  • • Public safety, such as construction management, traffic violation
Friendly Community
  • • Lost and found
  • • Looking for missing
  • • Locate patients' footprints
STAI Key Features

Real-time Video Analysis & Complex Event Detection

  • • Automate monitoring and alert
  • • Rule-based complex event processing (tagging/detection)
  • • AI-based object/action tagging/detection

Trace-Track and Foresee by AI-based Reasoning

  • • Automate reasoning out who and what objects/persons relevant
  • • Automate and continuously tracking and tracing the persons involved and visualize locations and paths

Exploratory Event-Based Analysis of Video

  • • Real-time dashboard and custom reporting for actionable decision-making
  • • Analyze and explore the trends and root causes

STAI運用多項自主研發的專利關鍵技術,在現有監控攝影機 (CCTV) 設備下,提供智能化影像推理與分析服務(SVRA),包含
▸ 複雜事件偵測(Complex Event Detection):隨時且快速地處理海量的監控攝影機(CCTV)影像數據 ,自動分析一連串行為與事件,即時偵測複雜事件的發生,協助場域管理者(如保安人員)可以在事件發生時即刻前往處理。
▸ 事件推理與追蹤(Track and Trace by AI-based Reasoning):利用人工智能推理,了解事件發生脈絡,自動追朔事件相關者,互動式搜索事件相關人並視覺化呈現移動足跡,以協助場域管理者(如保安人員)在最快時間內找到事件相關者所在位置
▸ 事件因果分析(Causality and Association Analysis):分析複雜事件數據,找出趨勢和根本原因,即時提供分析報表,做為場域管理與安全的決策支援。

For one single CCTV
Local Tracking
For overlapped view areas
across multiple CCTVs
Geo-location-based Re-Identification
For non-overlapped view areas
across multiple CCTVs
Feature-based Re-Identification
STAI Video Event Data Process Architecture

SVRA is an SaaS platform that provides service to help customers to collect, process and analyze sensory events that are derived from a vast area with multiple surveillance cameras and sensors.


  • ▸ Ultra-fast processing and computing video event data
  • ▸ 3 deployment models: on-premises, private cloud, public cloud
  • ▸ Support edge database
  • ▸ Cluster architecture: High-scalability and flexibility
  • ▸ Protect people's privacy and safety
  • ▸ Could integrate with biometric identification system, such as facial recognition
  • ▸ Use already existing CCTVs/IoT sensors without buying new devices
Contact Us