Pornography-800

该数据集包含800个视频，分为400个色情视频和400个非色情视频，用于基于音频的色情内容检测。此外，数据集还包括独特的音频特征，如呻吟声、深呼吸和背景音，经过人工检查，这两个类别的音频都有其特定的声音特征。规模上，该数据集包含800个音频样本（其中400个为色情内容，400个为非色情内容），其任务是对色情与非色情音频进行二分类。This dataset consists of 800 videos, split into 400 pornographic videos and 400 non-pornographic videos, intended for audio-based pornographic content detection. Additionally, the dataset includes unique audio features such as moans, deep breaths, and background sounds. After manual inspection, the audio samples from both categories have their specific acoustic characteristics. In terms of scale, the dataset contains 800 audio samples, with 400 corresponding to pornographic content and 400 to non-pornographic content, and its task is binary classification between pornographic and non-pornographic audio.