Dream2Image
收藏资源简介:
Dream2Image是一个结合了睡眠期间的EEG信号、梦境报告和AI生成的图像的开放多模态数据集。该数据集基于38名参与者和超过31小时的梦境EEG记录,包含129个样本,提供了醒前的最后几秒钟的大脑活动(T-15,T-30,T-60,T-120)、梦境经历的原始报告以及梦境的近似视觉重建。这个数据集为梦境研究提供了一个新的资源,一个独特的研究梦境神经相关性的资源,用于开发从大脑活动中解码梦境的模型,并在神经科学、心理学和人工智能领域探索新的方法。Dream2Image在Hugging Face和GitHub上公开提供,旨在支持在人工智能和神经科学的交界处进行的研究。
Dream2Image is an open multi-modal dataset that combines EEG signals during sleep, dream reports, and AI-generated images. This dataset is based on 38 participants and over 31 hours of dream EEG recordings, containing 129 samples. It provides brain activity recordings from the final seconds prior to awakening (T-15, T-30, T-60, T-120), original reports of dream experiences, and approximate visual reconstructions of dreams. This dataset constitutes a novel resource for dream research, and a unique resource for investigating the neural correlates of dreams, for developing models that decode dreams from brain activity, and for exploring innovative methodologies across the fields of neuroscience, psychology, and artificial intelligence. Dream2Image is publicly accessible via Hugging Face and GitHub, with the goal of supporting research at the intersection of artificial intelligence and neuroscience.

- 1Dream2Image : An Open Multimodal EEG Dataset for Decoding and Visualizing Dreams with Artificial Intelligence加利福尼亚大学圣地亚哥分校神经科学系 · 2025年



