ERIT Lightweight Multimodal Dataset for Elderly Emotion Recognition

<p><span>ERIT is a novel multimodal dataset designed to facilitate research in elderly emo</span><span>tion recognition. The dataset contains lightweight text and image data collected</span><br><span>from videos of elderly individuals reacting to various stimuli. The dataset includes&nbsp;</span><span>transcriptions of the speech and emotion labels extracted from the video frames. It was created for multimodal fusion evaluation, but can be also used for elderly emotion recognition from the image.<br></span></p>