MissBench: Benchmarking Multimodal Affective Analysis under Imbalanced Missing Modalities

TL;DR

MissBench introduces a standardized benchmark and diagnostic metrics for evaluating multimodal affective analysis models under realistic imbalanced missing data scenarios, revealing robustness gaps in existing methods.

Contribution

The paper presents MissBench, a benchmark with protocols and metrics for assessing multimodal affective models under imbalanced missing modalities, highlighting issues not captured by traditional evaluations.

Findings

Models can appear robust under shared missing rates but show inequity under imbalanced conditions.

The Modality Equity Index (MEI) reveals fairness issues among modalities.

The Modality Learning Index (MLI) uncovers optimization imbalances during training.

Abstract

Multimodal affective computing underpins key tasks such as sentiment analysis and emotion recognition. Standard evaluations, however, often assume that textual, acoustic, and visual modalities are equally available. In real applications, some modalities are systematically more fragile or expensive, creating imbalanced missing rates and training biases that task-level metrics alone do not reveal. We introduce MissBench, a benchmark and framework for multimodal affective tasks that standardizes both shared and imbalanced missing-rate protocols on four widely used sentiment and emotion datasets. MissBench also defines two diagnostic metrics. The Modality Equity Index (MEI) measures how fairly different modalities contribute across missing-modality configurations. The Modality Learning Index (MLI) quantifies optimization imbalance by comparing modality-specific gradient norms during training, aggregated across modality-related modules. Experiments on representative method families show that models that appear robust under shared missing rates can still exhibit marked modality inequity and optimization imbalance under imbalanced conditions. These findings position MissBench, together with MEI and MLI, as practical tools for stress-testing and analyzing multimodal affective models in realistic incomplete-modality settings.For reproducibility, we release our code at: https://anonymous.4open.science/r/MissBench-4098/