Interpreting vision and language generative models with semantic visual priors

Frontiers in Artificial Intelligence

Artificial Intelligence

Kees van Deemter¹  Michele Cafagna²  Albert Gatt³  Lina M. Rojas-Barahona⁴ 

[1] Department of Information and Computing Sciences, Utrecht University, Utrecht, Netherlands;Institute of Linguistics and Language Technology, University of Malta, Msida, Malta;Institute of Linguistics and Language Technology, University of Malta, Msida, Malta;Department of Information and Computing Sciences, Utrecht University, Utrecht, Netherlands;Orange Innovation, Lannion, France;

关键词: vision and language;    multimodality;    explainability;    image captioning;    visual question answering;    natural language generation;   

DOI  :  10.3389/frai.2023.1220476

 received in 2023-05-10, accepted in 2023-09-04,  发布年份 2023

来源: Frontiers

【 摘 要 】

When applied to Image-to-text models, explainability methods have two challenges. First, they often provide token-by-token explanations namely, they compute a visual explanation for each token of the generated sequence. This makes explanations expensive to compute and unable to comprehensively explain the model's output. Second, for models with visual inputs, explainability methods such as SHAP typically consider superpixels as features. Since superpixels do not correspond to semantically meaningful regions of an image, this makes explanations harder to interpret. We develop a framework based on SHAP, that allows for generating comprehensive, meaningful explanations leveraging the meaning representation of the output sequence as a whole. Moreover, by exploiting semantic priors in the visual backbone, we extract an arbitrary number of features that allows the efficient computation of Shapley values on large-scale models, generating at the same time highly meaningful visual explanations. We demonstrate that our method generates semantically more expressive explanations than traditional methods at a lower compute cost and that it can be generalized to a large family of vision-language models.

【 授权许可】

Unknown   
Copyright © 2023 Cafagna, Rojas-Barahona, van Deemter and Gatt.