New Product

Edge AI Station α

AI Station with bundled video AI application

About Edge AI Station α

Edge AI Station α is a “Video AI Application Bundled Edge AI Station” that provides an “All-in-One” AI application development and execution environment, pre-installed with a general-purpose AI application, as a set with Edge AI Station. Edge AI Station α reduces the burden of developing AI applications and AI models for use cases, enabling rapid deployment and cost reduction.

Upcoming Edge AI Station Alpha Releases

AI application development for Edge AI Station α by a development partner with extensive experience and achievements in AI application development

※The release of new models and the selection of AI applications will also be announced on this website.

The first EAIS α “Kaleidoscope-equipped model” is now available!

As the first release of Edge AI Station Alpha
The model bundled with “Kaleidoscope,” a multifunctional headcount application, is now available.

AI Application Overview

A single AI application that detects a person's whole body or head and performs the following functions

Counting the number of people in an area

Counting the number of people crossing a line

Intrusion detection within the area

Detection of presence within an area
※The above functions will be added and released when the AI application is upgraded.

Post-processing (Node-RED) Section Overview

Node-RED flow for generic external system integration will be provided

※Node-RED is a visual programming development tool for interconnecting hardware devices, APIs, and online services。

Feature implementation (release) schedule

※Newly implemented functions will be available by upgrading the AI application.

v1.0 (first release: March 2025)

Detection of people by the whole body

Three functions are implemented: counting the number of people in the area, counting the number of people crossing the line, and detecting intrusion into the area.

v1.x (to be released in 2025)

Implementing a feature to detect when someone is present in an area

v2.0 or later (to be released in 2025)

Implementing a human head detection model

All functions can be specified to detect either the whole body or the head of a person

Kaleidoscope Sample

For one pipeline (one video input stream), the functions of counting the number of people within an area and counting the number of people passing through are performed simultaneously.

The images of Yamashita Park@Yokohama taken by our company are applied to Kaleidoscope(v1.0).(※Under development sample)
Contact ▶