MECO: A Multimodal Dataset for Emotion and Cognitive Understanding in Older Adults
Summary: arXiv:2604.03050v1 Announce Type: cross
The field of affective computing has made significant advancements in recent years; however, the area of multimodal emotion prediction specifically within aging populations has remained largely unexplored. This gap is primarily due to the limited availability of datasets specifically designed for older adults. Most existing multimodal benchmarks focus on younger, cognitively healthy individuals and overlook the profound impact that cognitive decline can have on emotional expression and physiological responses. To address this critical issue, researchers have introduced MECO, a Multimodal dataset for Emotion and Cognitive understanding in Older adults.
MECO comprises data from 42 participants, providing approximately 38 hours of multimodal signals. This extensive dataset yields a total of 30,592 synchronized samples, making it a rich resource for researchers in the field. To enhance ecological validity, the data collection process adhered to standardized protocols within community-based settings, ensuring that the findings reflect real-world scenarios.
Dataset Composition
The MECO dataset encompasses a variety of modalities, which include:
- Video: Capturing non-verbal cues and visual emotions.
- Audio: Recording vocal expressions and tone variations.
- Electroencephalography (EEG): Measuring brain activity associated with emotional and cognitive states.
- Electrocardiography (ECG): Monitoring heart rate variability linked to emotional responses.
Emotional and Cognitive Annotations
In addition to the diverse modalities, the MECO dataset provides comprehensive annotations that include:
- Self-assessed valence: Participants’ evaluations of their emotional state.
- Arousal levels: The intensity of emotional reactions.
- Six basic emotions: Including happiness, sadness, anger, surprise, fear, and disgust.
- Mini-Mental State Examination (MMSE) scores: Assessing cognitive function and potential decline.
Baseline Benchmarks
To facilitate further research, the MECO dataset establishes baseline benchmarks for both emotional and cognitive prediction. These benchmarks serve as a reference point for future studies focusing on the complex interplay between emotion and cognition in older adults.
Applications and Future Directions
MECO is poised to serve as a foundational resource for multimodal modeling in the context of aging populations. Its implications extend to various applications, such as:
- Personalized emotion recognition systems tailored for older adults.
- Early detection of mild cognitive impairment (MCI) in real-world settings.
- Development of supportive technologies that improve the quality of life for aging individuals.
The complete MECO dataset and supplementary materials are accessible online, providing researchers with valuable tools to advance the understanding of emotional and cognitive dynamics in older populations. For more details, visit MECO Page.