오프라인 마케팅의 새로운 패러다임 - AI를 활용한 옥외광고효과측정

addd's Artificial Intelligence(AI) Technology

Pose Estimation

Real-time analysis of pose estimation for pedestrians(objects), extracting data such as gender, age, Exposed/Watching/Attention rates.

Object Detection

Built its own dataset, conducted advanced pose tracking, safety detection, etc., through lightweight edge devices.

Privacy protection

Privacy Protection

Real-time analysis and processing of pedestrians(objects) without anonymization, recording/storage, utilizing proxy servers and IP security for stable data protection, management, and transmission.

Additional Sensors Analysis

We collect wind speed, humidity, climate, and other environmental data through sensors, performing correlation analysis

Video collection

01 Vision sensor collection

AI analysis

02 Anonymization

03 Data A/R

Reporting

04 Provide analysis results

Approx 1 billion

GPS-based data

Approx 300 million

Mobility driving pattern data

Approx 900 million

Exposed pop. recognition data

addd's Collecting Data

Dataset collected so far

Approx 200 million

Point cloud data

Approx 7,000

Mobility analysis data

7 months

Pedestrian path analysis data

Approx 600 million

Environmental data

Web Dashboard

Implementing AI analysis data into a dashboard for display.

Try the demo

Aggregating the demographic data of the ad-exposed population

Processing 2 million pixel frames at a speed of 15 FPS,
analyzing a maximum of approximately 120 people per frame.

Aggregated Population Gender/Age Prediction

Predicting gender and age based on the attire and physical characteristics of each captured individual.

Aggregation of ad watching/attention population

Analyzing the angles (left/right, up/down, tilt) at which the heads of the exposed population are facing to determine whether they are watching/paying attention to the advertisement.

addd's Technology Details