DEVIATE
AI-powered platform for analyzing consumer behavior in physical retail stores
Detailed project description
The DEVIATE project focuses on transforming retail analytics through the integration of AI-based video analytics tools for real-time assessment of consumer behaviour.
Combining deep learning techniques, computer vision, and demographic modelling, the platform processes image and video data with the aim of extracting actionable insights about consumer profiles and their interaction with products on retail store shelves.
Using RGB cameras and advanced neural networks such as convolutional neural networks (CNNs) and recurrent neural networks (LSTMs), DEVIATE enables automatic detection of customer presence, tracking of movement, as well as estimation of gender and age within the store.
The data is used to optimize sales strategies, design store layouts, and determine product placement.
A key feature of the project is the use of scalable AI modules that operate in real time and are trained on large-scale video datasets from retail stores.
The system’s modular architecture allows easy integration into existing retail infrastructures, providing commercial stakeholders with a powerful tool for monitoring customer interaction, understanding preferences, and supporting data-driven decision-making.
The platform enhances operational performance while also supporting ethical analytics practices and privacy protection, contributing to the development of intelligent and adaptive retail environments.
Type and scope of work provided
- Development of a consumer behavior analysis system using deep learning
- Design of artificial intelligence models for real-time detection of presence, age, and gender
- Integration of computer vision technologies with video infrastructures and RGB camera systems
- Implementation of behavior tracking algorithms for analyzing interaction with products
- Development of a scalable analytics platform to support decision-making in retail
DEVIATE contributes to the digital transformation of the retail sector by providing tools for demographic analysis, behavioral pattern recognition, and real-time interaction analysis. The platform empowers businesses with predictive data to improve operational planning, product placement, and personalized marketing strategies.
The project is implemented under the National Recovery and Resilience Plan “Greece 2.0”, funded by the European Union – NextGenerationEU (Project code: ΥΠ3ΤΑ-0560460 – Budget: 1,800,000€).
Ψηφιακός Μετασχηματισμός Μικρομεσαίων Επιχειρήσεων
Ημερομηνία:
April 2025
