SER_AMPEL
收藏资源简介:
SER_AMPEL是一个专为意大利老年人设计的语音情感识别多源数据集,由米兰比可卡大学信息、系统和通信系创建。该数据集包含三个子集,分别收集自自然对话和电影、电视系列中的表演对话,旨在通过不同情境诱发情感。数据集的创建过程涉及对话的分割、标注和转录,应用领域包括人机交互和老年人护理,旨在通过识别情感状态提高交互的自然性和社会性。
SER_AMPEL is a multi-source speech emotion recognition dataset specifically designed for the elderly population in Italy, developed by the Department of Informatics, Systems and Communication of the University of Milan-Bicocca. This dataset includes three subsets, which are collected from both natural conversations and acted dialogues sourced from movies and television series, aiming to evoke emotions across different contexts. The dataset creation process involves dialogue segmentation, annotation and transcription. Its application fields cover human-computer interaction and elder care, with the goal of improving the naturalness and sociality of interactions by recognizing emotional states.




