The sound of crashing waves, the roar of fast-moving cars -- sound conveys important information about the objects in our surroundings. In this work, we show that ambient sounds can be used as a supervisory signal for learning visual models. To demonstrate this, we train a convolutional neural network to predict a statistical summary of the sound associated with a video frame. We show that, through this process, the network learns a representation that conveys information about objects and scenes. We evaluate this representation on several recognition tasks, finding that its performance is comparable to that of other state-of-the-art unsupervised learning methods. Finally, we show through visualizations that the network learns units that are selective to objects that are often associated with characteristic sounds.
@article{arxiv.1608.07017,
title = {Ambient Sound Provides Supervision for Visual Learning},
author = {Andrew Owens and Jiajun Wu and Josh H. McDermott and William T. Freeman and Antonio Torralba},
journal= {arXiv preprint arXiv:1608.07017},
year = {2016}
}