Learning a Predictable and Generative Vector Representation for Objects

In European Conference on Computer Vision, 2016 (Spotlight Presentation)

Rohit Girdhar, David F. Fouhey, Mikel Rodriguez, Abhinav Gupta

Teaser figure for Learning a Predictable and Generative Vector Representation for Objects

Abstract

What is a good vector representation of an object? We believe that it should be generative in 3D, in the sense that it can produce new 3D objects; as well as be predictable from 2D, in the sense that it can be perceived from 2D images. We propose a novel architecture, called the TL-embedding network, to learn an embedding space with these properties. The network consists of two components: (a) an autoencoder that ensures the representation is generative; and (b) a convolutional network that ensures the representation is predictable. This enables tackling a number of tasks including voxel prediction from 2D images and 3D model retrieval. Extensive experimental analysis demonstrates the usefulness and versatility of this embedding.

Citation

@inproceedings{girdhar2016learning,
    title = {Learning a Predictable and Generative Vector Representation for Objects},
    author = {Girdhar, Rohit and Fouhey, David F. and Rodriguez, Mikel and Gupta, Abhinav},
    booktitle = {ECCV},
    year = {2016},
}