Pet Vocalization Emotion Annotation Dataset: Public Preview

This record describes a limited public preview of Thordata DataMall's Pet Vocalization Emotion Annotation Dataset. The underlying dataset contains cat and dog vocalization recordings with bilingual Chinese-English emotion and behavioral-state annotations. Audio clips are short MP3 segments of approximately 5–20 seconds and include basic audio metadata such as filename, format, duration, sampling rate, channel count, and, where available, bitrate. The dataset is intended for exploratory research in animal vocalization classification, emotion recognition, audio retrieval, sound-event analysis, and multimodal AI prototyping. This record contains a metadata manifest and links to nine DataMall-hosted preview samples. It does not redistribute the audio files or the complete commercial dataset. Usage, redistribution, model training, fine-tuning, and commercial permissions remain subject to the rights holder's approved sample-release terms. The license status is pending confirmation at the time of preparation. Source product:https://datamall.thordata.com/datasets/AUD-20260910-151824-274