Evaluating the Representational Hub of Language and Vision Models
This work addresses the challenge of optimizing multimodal representations for AI systems, though it is incremental as it builds on existing cognitive science architectures.
The paper investigates how different vision-and-language tasks affect the encoder in multimodal models, finding that tasks like visual question answering improve performance on diagnostic tests by up to 15% compared to baseline models.
The multimodal models used in the emerging field at the intersection of computational linguistics and computer vision implement the bottom-up processing of the `Hub and Spoke' architecture proposed in cognitive science to represent how the brain processes and combines multi-sensory inputs. In particular, the Hub is implemented as a neural network encoder. We investigate the effect on this encoder of various vision-and-language tasks proposed in the literature: visual question answering, visual reference resolution, and visually grounded dialogue. To measure the quality of the representations learned by the encoder, we use two kinds of analyses. First, we evaluate the encoder pre-trained on the different vision-and-language tasks on an existing diagnostic task designed to assess multimodal semantic understanding. Second, we carry out a battery of analyses aimed at studying how the encoder merges and exploits the two modalities.