Okapi: Generalising Better by Making Statistical Matches Match

TL;DR

Okapi introduces a robust semi-supervised learning method using online statistical matching with a nearest-neighbours approach, improving out-of-distribution generalization across diverse modalities and tasks.

Contribution

The paper presents a novel, efficient semi-supervised learning technique based on statistical matching that is modality- and task-agnostic, outperforming existing methods on real-world datasets.

Findings

Outperforms baseline methods in OOD generalization on multiple datasets

Leverages unlabelled data to improve over empirical risk minimization

Produces semantically meaningful matches in learned representations

Abstract

We propose Okapi, a simple, efficient, and general method for robust semi-supervised learning based on online statistical matching. Our method uses a nearest-neighbours-based matching procedure to generate cross-domain views for a consistency loss, while eliminating statistical outliers. In order to perform the online matching in a runtime- and memory-efficient way, we draw upon the self-supervised literature and combine a memory bank with a slow-moving momentum encoder. The consistency loss is applied within the feature space, rather than on the predictive distribution, making the method agnostic to both the modality and the task in question. We experiment on the WILDS 2.0 datasets Sagawa et al., which significantly expands the range of modalities, applications, and shifts available for studying and benchmarking real-world unsupervised adaptation. Contrary to Sagawa et al., we show that it is in fact possible to leverage additional unlabelled data to improve upon empirical risk minimisation (ERM) results with the right method. Our method outperforms the baseline methods in terms of out-of-distribution (OOD) generalisation on the iWildCam (a multi-class classification task) and PovertyMap (a regression task) image datasets as well as the CivilComments (a binary classification task) text dataset. Furthermore, from a qualitative perspective, we show the matches obtained from the learned encoder are strongly semantically related. Code for our paper is publicly available at https://github.com/wearepal/okapi/.