Recursive Neural Programs: Variational Learning of Image Grammars and Part-Whole Hierarchies

Ares Fisher, Rajesh P. N. Rao

Keywords: computer vision, generative models, composing representations, image grammar

TL;DR: We introduce a neural generative model that addresses the part-whole hierarchy learning problem by modeling images as recursive hierarchical trees of probabilistic sensory-motor programs, enabling intuitive composition and learning of image grammars.

Abstract: Human vision involves parsing and representing objects and scenes using structured representations based on part-whole hierarchies. Computer vision and machine learning researchers have recently sought to emulate this capability using capsule networks, object reference frames and active predictive coding, but a generative model formulation has been lacking. We introduce Recursive Neural Programs (RNPs), a neural generative model that addresses the part-whole hierarchy learning problem by modeling images as hierarchical trees of probabilistic sensory-motor programs. These programs recursively reuse learned sensory-motor primitives to model an image within different reference frames, enabling intuitive and explainable composition and allowing for forming recursive image grammars. We express RNPs as structured variational autoencoders (sVAEs) for inference and sampling, and demonstrate parts-based parsing, sampling and one-shot transfer learning for MNIST, Omniglot and ETH-80 datasets. Our results show that RNPs provide an intuitive and explainable way of composing objects and scenes, allowing rich compositionality and intuitive interpretations of objects in terms of part-whole hierarchies.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Generative models

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/recursive-neural-programs-variational/code)