DeepAI AI Chat
Log In Sign Up

Ambient Sound Provides Supervision for Visual Learning

by   Andrew Owens, et al.

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.


page 2

page 4

page 8

page 19

page 20


Learning Sight from Sound: Ambient Sound Provides Supervision for Visual Learning

The sound of crashing waves, the roar of fast-moving cars -- sound conve...

Local Visual Microphones: Improved Sound Extraction from Silent Video

Sound waves cause small vibrations in nearby objects. A few techniques e...

Improving Visual Recognition using Ambient Sound for Supervision

Our brains combine vision and hearing to create a more elaborate interpr...

Structure from Silence: Learning Scene Structure from Ambient Sound

From whirling ceiling fans to ticking clocks, the sounds that we hear su...

Learning to Localize Sound Source in Visual Scenes

Visual events are usually accompanied by sounds in our daily lives. We p...

SoundNet: Learning Sound Representations from Unlabeled Video

We learn rich natural sound representations by capitalizing on large amo...

Impulsive Sound Detection by a Novel Energy Formula and its Usage for Gunshot Recognition

There are many methods proposed for the detection of impulsive sounds in...