Designing for Sustainable Choices: Leveraging Computer Vision and AI to Analyze Eco-Friendly Product Labels

Consumers increasingly seek sustainable products, yet the effectiveness of eco-labels in influencing purchasing decisions varies widely. This project will focus on the initial stages of the Design Thinking process—Discover and Define—to study how visual cues on product packaging related to sustainability affect consumer trust and choice. The insights will guide the creation of tools to improve label transparency and encourage greener consumer behavior.

1. Computer Vision Analysis:
   Apply computer vision techniques such as image classification and object detection to a large dataset of product packaging images. The goal is to identify and categorize sustainability-related visual elements—eco-labels, certification logos, ingredient icons, and color schemes—to map their prevalence and diversity across product categories.
2. Data-Driven Consumer Segmentation:
   Using machine learning clustering and AI-assisted narrative generation, develop consumer segments based on how different demographic groups respond to sustainability cues. This will involve prompt engineering with Generative AI to create realistic user personas representing eco-conscious shoppers, skeptical buyers, and indifferent consumers.

• Implement image processing workflows to extract and classify eco-label features from packaging images.
• Perform clustering analyses to segment consumers by response patterns to sustainability labels.
• Use Generative AI to craft detailed user narratives and visual personas for each segment.
• Compile results in a final report and deliver a presentation highlighting methods, findings, and design implications.