MELDAE: A Framework for Micro-Expression Spotting, Detection, and Automatic Evaluation in In-the-Wild Conversational Scenes

TL;DR

This paper introduces MELDAE, a comprehensive framework for micro-expression spotting and detection in natural conversational scenes, featuring a new dataset, an end-to-end model, and a boundary-aware loss for improved temporal accuracy.

Contribution

The paper presents the first in-the-wild conversational micro-expression dataset, an end-to-end detection framework, and a boundary-aware loss function to enhance temporal localization accuracy.

Findings

Achieved state-of-the-art results on WDMD dataset with a 17.72% improvement in F1_{DR} localization metric.

Demonstrated strong generalization on existing benchmarks.

Proposed boundary-aware loss improves temporal detection precision.

Abstract

Accurately analyzing spontaneous, unconscious micro-expressions is crucial for revealing true human emotions, but this task remains challenging in wild scenarios, such as natural conversation. Existing research largely relies on datasets from controlled laboratory environments, and their performance degrades dramatically in the real world. To address this issue, we propose three contributions: the first micro-expression dataset focused on conversational-in-the-wild scenarios; an end-to-end localization and detection framework, MELDAE; and a novel boundary-aware loss function that improves temporal accuracy by penalizing onset and offset errors. Extensive experiments demonstrate that our framework achieves state-of-the-art results on the WDMD dataset, improving the key F1_{DR} localization metric by 17.72% over the strongest baseline, while also demonstrating excellent generalization capabilities on existing benchmarks.