MEWL: Few-shot multimodal word learning with referential uncertainty

TL;DR

This paper introduces MEWL, a benchmark suite for evaluating how well machines can learn word meanings in grounded visual scenes with minimal examples, inspired by human cognitive abilities.

Contribution

The paper presents a systematic benchmark for human-like few-shot multimodal word learning, filling a gap in evaluating machines' grounded language understanding.

Findings

Humans outperform machines in few-shot word learning tasks.

Current models show a significant gap compared to human performance.

The benchmark aligns with developmental theories of language acquisition.

Abstract

Without explicit feedback, humans can rapidly learn the meaning of words. Children can acquire a new word after just a few passive exposures, a process known as fast mapping. This word learning capability is believed to be the most fundamental building block of multimodal understanding and reasoning. Despite recent advancements in multimodal learning, a systematic and rigorous evaluation is still missing for human-like word learning in machines. To fill in this gap, we introduce the MachinE Word Learning (MEWL) benchmark to assess how machines learn word meaning in grounded visual scenes. MEWL covers human's core cognitive toolkits in word learning: cross-situational reasoning, bootstrapping, and pragmatic learning. Specifically, MEWL is a few-shot benchmark suite consisting of nine tasks for probing various word learning capabilities. These tasks are carefully designed to be aligned with the children's core abilities in word learning and echo the theories in the developmental literature. By evaluating multimodal and unimodal agents' performance with a comparative analysis of human performance, we notice a sharp divergence in human and machine word learning. We further discuss these differences between humans and machines and call for human-like few-shot word learning in machines.