Convolutional Conditional Neural Processes
This work addresses the need for translation-equivariant models in machine learning, offering a novel approach that improves performance and generalization for various data types, though it builds incrementally on the Neural Process family.
The paper tackled the problem of modeling translation equivariance in data for tasks like time series, spatial data, and images by introducing the Convolutional Conditional Neural Process (ConvCNP), which embeds data into an infinite-dimensional function space and achieves state-of-the-art performance with zero-shot generalization to out-of-domain tasks.
We introduce the Convolutional Conditional Neural Process (ConvCNP), a new member of the Neural Process family that models translation equivariance in the data. Translation equivariance is an important inductive bias for many learning problems including time series modelling, spatial data, and images. The model embeds data sets into an infinite-dimensional function space as opposed to a finite-dimensional vector space. To formalize this notion, we extend the theory of neural representations of sets to include functional representations, and demonstrate that any translation-equivariant embedding can be represented using a convolutional deep set. We evaluate ConvCNPs in several settings, demonstrating that they achieve state-of-the-art performance compared to existing NPs. We demonstrate that building in translation equivariance enables zero-shot generalization to challenging, out-of-domain tasks.