데이터셋 탐색

Consists of more than 210k videos for 310 audio classes. Source: VGGSound: A Large-scale Audio-Visual Dataset

VGG-SS (VGG Sound Source) is a benchmark for evaluating sound source localisation in videos. The dataset consists on a new set of annotations for the recently-introduced VGG-Sound dataset, where the sound sources visible in each video clip are explicitly marked with bounding box annotations. This dataset is 20 times larger than analogous existing ones, contains 5K videos spanning over 200 categories, and, differently from Flickr SoundNet, is video-based.

VGG-Sound Sync is an audio-visual synchronisation benchmark based on videos collected from YouTube. VGG-Sound Sync contains over 100k video clips, spanning 160 classes and can be downloaded here. Note, only the test clips are included here, please use the training clips in the original VGG-Sound to train your models ( classes are same with the ones in the test clips). Each line in the json file has been defined by: # YouTube ID, start seconds, label