MuDD: A Multimodal Deception Detection Dataset and GSR-Guided Progressive Distillation for Non-Contact Deception Detection

Abstract

Non-contact automatic deception detection remains challenging because visual and auditory deception cues often lack stable cross-subject patterns. In contrast, galvanic skin response (GSR) provides more reliable physiological cues and has been widely used in contact-based deception detection. In this work, we leverage stable deception-related knowledge in GSR to guide representation learning in non-contact modalities through cross-modal knowledge distillation. A key obstacle, however, is the lack of a suitable dataset for this setting. To address this, we introduce MuDD, a large-scale Multimodal Deception Detection dataset containing recordings from 130 participants over 690 minutes. In addition to video, audio, and GSR, MuDD also provides Photoplethysmography, heart rate, and personality traits, supporting broader scientific studies of deception. Based on this dataset, we propose GSR-guided Progressive Distillation (GPD), a cross-modal distillation framework for mitigating the negative transfer caused by the large modality mismatch between GSR and non-contact signals. The core innovation of GPD is the integration of progressive feature-level and digit-level distillation with dynamic routing, which allows the model to adaptively determine how teacher knowledge should be transferred during training, leading to more stable cross-modal knowledge transfer. Extensive experiments and visualizations show that GPD outperforms existing methods and achieves state-of-the-art performance on both deception detection and concealed-digit identification.

Figures (4)

• Figure 1: Motivation of this work. (a) Visual deception cues are often weak, subtle, and subject-specific, making it difficult to identify stable cross-subject patterns, whereas GSR exhibits more consistent deception-related responses across individuals. (b) We illustrate two possible directions for improving non-contact deception detection: scaling up data collection and annotation, or transferring more stable discriminative knowledge from GSR to non-contact modalities via knowledge distillation. This work investigates the latter direction, especially under limited-data conditions.
• Figure 2: Overview of the proposed GSR-guided Progressive Distillation (GPD) framework.
• Figure 3: Prediction-transition statistics for the video and audio modalities. Samples are grouped by student prediction changes before and after distillation, and further split by whether the teacher prediction is correct or incorrect.
• Figure 4: Representative JSD-based response curves for the video and audio modalities, comparing the teacher and the student before and after distillation.