iStructTab: Structured Feature Sequencing for Multimodal Learning of Image and Tabular Data

2026-08-04

AI Summary

This work addresses the inefficiencies in multimodal fusion of images and tabular data—specifically, feature redundancy, dispersion, and limited generalization arising from suboptimal feature representations—by proposing a structured feature ranking approach. The method uniquely integrates Graph-Enhanced Descriptor Sorting (GEDS) with a sequence-aware efficient Transformer: it first constructs similarity graphs to generate statistical descriptors and optimizes column permutations, then explicitly models the structural ordering of tabular features within the Transformer using sequence-aware memory tokens and a dedicated ranking loss function. Experimental results across multiple multimodal benchmarks demonstrate that the proposed approach significantly reduces feature dispersion, enhances predictive performance, and improves model robustness, thereby validating the efficacy of structured feature ranking in multimodal fusion.

Abstract

Multimodal learning of images and tabular data is often impaired by ineffective representations, resulting in redundancy, dispersion, and generalization problems. To tackle this challenge, we introduce Graph-Enhanced Descriptor Sequencing (GEDS), a structured feature sequencing algorithm grounded in principles from the Column Permutation Problem (CPP). GEDS refines statistical descriptors of the features through similarity graph-based computations, systematically determining an effective feature sequencing. We incorporate GEDS within an order-aware efficient transformer framework, utilizing order-aware memory tokens that explicitly adhere to the derived feature sequencing via a dedicated loss function. Experimental results across multimodal benchmarks demonstrate that iStructTab effectively minimizes feature dispersion, improving predictive performance and robustness, and highlighting the significance of structured feature sequencing in multimodal learning.

Problem

• multimodal learning
• feature representation
• image and tabular data
• feature dispersion
• generalization

Innovation

• structured feature sequencing
• multimodal learning
• graph-enhanced descriptor
• order-aware transformer
• feature dispersion