Energy-Based Processes for Exchangeable Data
arXiv:2003.07521
Abstract
Recently there has been growing interest in modeling sets with exchangeability such as point clouds. A shortcoming of current approaches is that they restrict the cardinality of the sets considered or can only express limited forms of distribution over unobserved data. To overcome these limitations, we introduce Energy-Based Processes (EBPs), which extend energy based models to exchangeable data while allowing neural network parameterizations of the energy function. A key advantage of these models is the ability to express more flexible distributions over sets without restricting their cardinality. We develop an efficient training procedure for EBPs that demonstrates state-of-the-art performance on a variety of tasks such as point cloud generation, classification, denoising, and image completion.
References in corpus (21)
- Stochastic Backpropagation and Approximate Inference in Deep Generative Models
- Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modeling
- Supervised Topic Models
- Structure Discovery in Nonparametric Regression through Compositional Kernel Search
- Deep Gaussian Processes
- Determinantal point process models and statistical inference : Extended version
- A Theory of Generative ConvNet
- Towards a Neural Statistician
- Student-t Processes as Alternatives to Gaussian Processes
- Deep Gaussian Processes for Regression using Approximate Expectation Propagation
- Stochastic Variational Deep Kernel Learning
- Implicit Generation and Generalization in Energy-Based Models
- PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows
- Conditional Neural Processes
- Multiresolution Tree Networks for 3D Point Cloud Processing
- The Functional Neural Process
- Attentive Neural Processes
- PointGrow: Autoregressively Learned Point Cloud Generation with Self-Attention
- Exponential Family Estimation via Adversarial Dynamics Embedding
- Neural Similarity Learning
- Variational Implicit Processes