Gentai-B012/IEMOCAP

该数据集是一个多模态情感分析数据集，包含音频文件及其相关特征和标注。数据集中每个样本包括音频文件路径、音频数据、多种情感强度（如沮丧、愤怒、悲伤、厌恶、兴奋、恐惧、中性、惊讶、快乐）的连续值评分、情感维度（激活度、效价、优势度）、说话者性别、转录文本、主要情感类别以及音频特征（如语速、音高均值、音高标准差、均方根振幅、相对分贝）。数据集分为训练集，共10,039个样本，适用于情感识别、语音处理和多模态机器学习任务。This dataset is a multimodal emotion analysis dataset containing audio files along with associated features and annotations. Each sample in the dataset includes the audio file path, audio data, continuous intensity ratings for multiple emotions (such as frustrated, angry, sad, disgust, excited, fear, neutral, surprise, happy), emotion dimensions (activation, valence, dominance), speaker gender, transcription text, major emotion category, and audio features (e.g., speaking rate, mean pitch, pitch standard deviation, root mean square amplitude, relative decibels). The dataset is split into a training set with 10,039 samples and is suitable for emotion recognition, speech processing, and multimodal machine learning tasks.