CompGuessWhat?!: A Multi-task Evaluation Framework for Grounded Language Learning

TL;DR

This paper introduces GROLLA, a comprehensive evaluation framework with three sub-tasks for assessing grounded language learning models, and presents the CompGuessWhat?! dataset to evaluate attribute grounding and generalization in neural representations.

Contribution

It proposes a multi-task evaluation framework and a new dataset to better assess the quality and generalization of grounded language learning models.

Findings

Current models have limited ability to encode object attributes (average F1 44.27).

Models struggle with zero-shot generalization, achieving only 50.06% accuracy.

The framework highlights the need for more expressive and robust representations.

Abstract

Approaches to Grounded Language Learning typically focus on a single task-based final performance measure that may not depend on desirable properties of the learned hidden representations, such as their ability to predict salient attributes or to generalise to unseen situations. To remedy this, we present GROLLA, an evaluation framework for Grounded Language Learning with Attributes with three sub-tasks: 1) Goal-oriented evaluation; 2) Object attribute prediction evaluation; and 3) Zero-shot evaluation. We also propose a new dataset CompGuessWhat?! as an instance of this framework for evaluating the quality of learned neural representations, in particular concerning attribute grounding. To this end, we extend the original GuessWhat?! dataset by including a semantic layer on top of the perceptual one. Specifically, we enrich the VisualGenome scene graphs associated with the GuessWhat?! images with abstract and situated attributes. By using diagnostic classifiers, we show that current models learn representations that are not expressive enough to encode object attributes (average F1 of 44.27). In addition, they do not learn strategies nor representations that are robust enough to perform well when novel scenes or objects are involved in gameplay (zero-shot best accuracy 50.06%).