What Do You See in Common? Learning Hierarchical Prototypes over Tree-of-Life to Discover Evolutionary Traits

TL;DR

This paper introduces HComP-Net, a hierarchical prototype learning framework that discovers evolutionary traits from biological images by modeling a tree structure, overcoming limitations of flat prototype methods.

Contribution

HComP-Net presents novel loss functions and a masking module to effectively learn hierarchical prototypes aligned with phylogenetic trees, improving trait discovery.

Findings

Prototypes are accurate and semantically consistent.

Model generalizes well to unseen species.

Outperforms baseline methods in hierarchical trait discovery.

Abstract

A grand challenge in biology is to discover evolutionary traits - features of organisms common to a group of species with a shared ancestor in the tree of life (also referred to as phylogenetic tree). With the growing availability of image repositories in biology, there is a tremendous opportunity to discover evolutionary traits directly from images in the form of a hierarchy of prototypes. However, current prototype-based methods are mostly designed to operate over a flat structure of classes and face several challenges in discovering hierarchical prototypes, including the issue of learning over-specific prototypes at internal nodes. To overcome these challenges, we introduce the framework of Hierarchy aligned Commonality through Prototypical Networks (HComP-Net). The key novelties in HComP-Net include a novel over-specificity loss to avoid learning over-specific prototypes, a novel discriminative loss to ensure prototypes at an internal node are absent in the contrasting set of species with different ancestry, and a novel masking module to allow for the exclusion of over-specific prototypes at higher levels of the tree without hampering classification performance. We empirically show that HComP-Net learns prototypes that are accurate, semantically consistent, and generalizable to unseen species in comparison to baselines.