Living Room TV Content Audio Background Sound Classification Dataset

In modern households, smart home devices need to recognize and adapt to the surrounding sound environment, but due to the diversity of audio signals, devices still face challenges in accuracy and real-time performance in background sound recognition. Existing audio recognition systems lack detailed categorization of background sounds and handling of special scenarios, making it difficult to respond accurately in changing environments. This dataset aims to address the classification of audio background sounds in living room TV content to meet the recognition needs of smart audio devices. Data was collected in real household environments using high-sensitivity microphone equipment to record audio data at different times and occasions. Quality control includes multiple rounds of annotation and consistency checks to ensure the accuracy of background sound classification. The annotation team consists of ten experts with experience in audio sorting. Data preprocessing employs advanced filtering and normalization techniques, and is organized and stored in WAV format.