Search papers, labs, and topics across Lattice.
The paper introduces FEAST, a retrieval-augmented framework for multi-hierarchical food classification within the FoodEx2 system, addressing challenges like complex label dependencies and data sparsity. FEAST decomposes the classification into base term identification, multi-label facet prediction, and facet descriptor assignment, leveraging the hierarchical structure and deep metric learning to create discriminative embeddings. Experiments on the multilingual FoodEx2 benchmark demonstrate that FEAST outperforms the CNN baseline, achieving F1 score improvements of 12-38% on rare classes.
Forget CNNs: a retrieval-augmented approach to food classification boosts F1 scores by up to 38% on rare classes within the complex FoodEx2 system.
Hierarchical text classification (HTC) and extreme multi-label classification (XML) tasks face compounded challenges from complex label interdependencies, data sparsity, and extreme output dimensions. These challenges are exemplified in the European Food Safety Authority's FoodEx2 system-a standardized food classification framework essential for food consumption monitoring and contaminant exposure assessment across Europe. FoodEx2 coding transforms natural language food descriptions into a set of codes from multiple standardized hierarchies, but faces implementation barriers due to its complex structure. Given a food description (e.g., "organic yogurt''), the system identifies its base term ("yogurt''), all the applicable facet categories (e.g., "production method''), and then, every relevant facet descriptors to each category (e.g., "organic production''). While existing models perform adequately on well-balanced and semantically dense hierarchies, no work has been applied on the practical constraints imposed by the FoodEx2 system. The limited literature addressing such real-world scenarios further compounds these challenges. We propose FEAST (Food Embedding And Semantic Taxonomy), a novel retrieval-augmented framework that decomposes FoodEx2 classification into a three-stage approach: (1) base term identification, (2) multi-label facet prediction, and (3) facet descriptor assignment. By leveraging the system's hierarchical structure to guide training and performing deep metric learning, FEASTlearns discriminative embeddings that mitigate data sparsity and improve generalization on rare and fine-grained labels. Evaluated on the multilingual FoodEx2 benchmark, FEAST outperforms the prior European's CNN baseline F1 scores by 12-38 % on rare classes.
