Learning to Generate Chairs, Tables and Cars with Convolutional Networks
arXiv:1411.5928
Abstract
We train generative 'up-convolutional' neural networks which are able to generate images of objects given object style, viewpoint, and color. We train the networks on rendered 3D models of chairs, tables, and cars. Our experiments show that the networks do not merely learn all images by heart, but rather find a meaningful representation of 3D models allowing them to assess the similarity of different models, interpolate between given views to generate the missing ones, extrapolate views, and invent new objects not present in the training set by recombining training instances, or even two different object classes. Moreover, we show that such generative networks can be used to find correspondences between different objects from the dataset, outperforming existing approaches on this task.
arXiv commentv4: final PAMI version. New architecture figure
References in corpus (9)
- Auto-Encoding Variational Bayes
- Caffe: Convolutional Architecture for Fast Feature Embedding
- DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition
- Depth Map Prediction from a Single Image using a Multi-Scale Deep Network
- Semi-Supervised Learning with Deep Generative Models
- Deep Convolutional Inverse Graphics Network
- Deep Generative Stochastic Networks Trainable by Backprop
- Random Walk Initialization for Training Very Deep Feedforward Networks
- Mixtures of conditional Gaussian scale mixtures applied to multiscale image representations
Cited by in corpus (17)
- Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
- Geometric deep learning: going beyond Euclidean data
- Deep Generative Image Models using a Laplacian Pyramid of Adversarial Networks
- Variational Autoencoder for Deep Learning of Images, Labels and Captions
- Generative Adversarial Active Learning
- Photographic Image Synthesis with Cascaded Refinement Networks
- GeneGAN: Learning Object Transfiguration and Attribute Subspace from Unpaired Data
- Multi-View Image Generation from a Single-View
- Conditional Generative Moment-Matching Networks
- Temporal Generative Adversarial Nets with Singular Value Clipping
- Towards Open-Set Identity Preserving Face Synthesis
- Associative Adversarial Networks
- Disentangled Representations in Neural Models
- Cross-Entropy Adversarial View Adaptation for Person Re-identification
- Unpaired Multi-Domain Image Generation via Regularized Conditional GANs
- Multiple Kernel Learning and Automatic Subspace Relevance Determination for High-dimensional Neuroimaging Data
- Multilinear Map Layer: Prediction Regularization by Structural Constraint