Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Personality Trait Recognition using ECG Spectrograms and Deep Learning

Muhammad Mohsin Altaf, Saadat Ullah Khan, Muhammad Majd, Syed Muhammad Anwar

TL;DR

This study addresses automatic recognition of Big Five personality traits from ECG signals by transforming ECG into spectrogram representations and applying deep learning models. It constructs a processing pipeline that generates 224×224 spectrograms via STFT with optimized window parameters, and leverages pretrained ResNet-18 and Vision Transformer architectures for binary trait classification on the ASCERTAIN dataset. The work demonstrates consistently high F1-scores (>0.9) across traits for multiple window settings, with ResNet-18 delivering robust performance while ViT shows potential given more data. Overall, the findings suggest ECG-based, objective personality assessment is viable and could augment traditional questionnaires for applications in personalized interfaces and healthcare.

Abstract

This paper presents an innovative approach to recognizing personality traits using deep learning (DL) methods applied to electrocardiogram (ECG) signals. Within the framework of detecting the big five personality traits model encompassing extra-version, neuroticism, agreeableness, conscientiousness, and openness, the research explores the potential of ECG-derived spectrograms as informative features. Optimal window sizes for spectrogram generation are determined, and a convolutional neural network (CNN), specifically Resnet-18, and visual transformer (ViT) are employed for feature extraction and personality trait classification. The study utilizes the publicly available ASCERTAIN dataset, which comprises various physiological signals, including ECG recordings, collected from 58 participants during the presentation of video stimuli categorized by valence and arousal levels. The outcomes of this study demonstrate noteworthy performance in personality trait classification, consistently achieving F1-scores exceeding 0.9 across different window sizes and personality traits. These results emphasize the viability of ECG signal spectrograms as a valuable modality for personality trait recognition, with Resnet-18 exhibiting effectiveness in discerning distinct personality traits.

Personality Trait Recognition using ECG Spectrograms and Deep Learning

TL;DR

This study addresses automatic recognition of Big Five personality traits from ECG signals by transforming ECG into spectrogram representations and applying deep learning models. It constructs a processing pipeline that generates 224×224 spectrograms via STFT with optimized window parameters, and leverages pretrained ResNet-18 and Vision Transformer architectures for binary trait classification on the ASCERTAIN dataset. The work demonstrates consistently high F1-scores (>0.9) across traits for multiple window settings, with ResNet-18 delivering robust performance while ViT shows potential given more data. Overall, the findings suggest ECG-based, objective personality assessment is viable and could augment traditional questionnaires for applications in personalized interfaces and healthcare.

Abstract

This paper presents an innovative approach to recognizing personality traits using deep learning (DL) methods applied to electrocardiogram (ECG) signals. Within the framework of detecting the big five personality traits model encompassing extra-version, neuroticism, agreeableness, conscientiousness, and openness, the research explores the potential of ECG-derived spectrograms as informative features. Optimal window sizes for spectrogram generation are determined, and a convolutional neural network (CNN), specifically Resnet-18, and visual transformer (ViT) are employed for feature extraction and personality trait classification. The study utilizes the publicly available ASCERTAIN dataset, which comprises various physiological signals, including ECG recordings, collected from 58 participants during the presentation of video stimuli categorized by valence and arousal levels. The outcomes of this study demonstrate noteworthy performance in personality trait classification, consistently achieving F1-scores exceeding 0.9 across different window sizes and personality traits. These results emphasize the viability of ECG signal spectrograms as a valuable modality for personality trait recognition, with Resnet-18 exhibiting effectiveness in discerning distinct personality traits.
Paper Structure (7 sections, 1 equation, 2 figures, 2 tables)

This paper contains 7 sections, 1 equation, 2 figures, 2 tables.

Table of Contents

  1. INTRODUCTION
  2. PROPOSED METHODOLOGY
  3. Electrocardiogram Data
  4. Pre-Processing and Spectrogram Generation
  5. Pre-Trained Deep Learning Models
  6. EXPERIMENTAL RESULTS AND DISCUSSION
  7. CONCLUSION

Figures (2)

  • Figure 1: Our proposed approach for personality recognition using ECG spectrograms and deep learning models.
  • Figure 2: Spectrograms of personality traits (a) Extrovert (b) Introvert (c) Agreeable (d) Non-Agreeable (e) Conscientiousness (f) Non-conscientiousness (g) Emotionally Stable (h) Emotionally Unstable (i) Open-minded (j) Close-minded.