Artificial intelligence for flavor perception: Integrating olfactory mechanisms into food group sensory evaluation

Abstract

Background

Flavor constitutes a pivotal determinant of food quality and sensory experience, with its evaluation conventionally relying on assessors and instruments. However, traditional approaches encounter significant challenges in terms of objectivity, efficiency, economic costs and throughput. In recent years, artificial intelligence (AI) offers novel methodologies and solutions to these bottlenecks.

Scope and approach

This paper presents a systematic review of AI-driven models for food group sensory evaluation, with particular emphasis on integrating flavor perception mechanisms, especially olfaction, into group sensory analysis. Furthermore, drawing from interdisciplinary perspectives including perceptual science, computational intelligence, and flavor-omics, it conducts an in-depth analysis of constructing models and its practical implementation. Particularly, grounded in the principles of olfactory perception, the review investigates recognition mechanism of flavor compounds and explores strategies to AI-driven flavor prediction, quality grading, and mechanistic interpretation.

Key findings and conclusions

Compared with traditional methods, AI-driven olfactory perception evaluation methods have greatly improved the analysis of differences within and between groups, efficiency, insight into flavor substances, and analysis of high-throughput and high-modal data. In addition, combined with specific algorithms and models, interpretability analysis and flavor prediction can be realized on the basis of sensory evaluation, holding strong promise for broader adoption in food science.