TY  - JOUR
AU  - Lluis Gomez
AU  - Ali Furkan Biten
AU  - Ruben Tito
AU  - Andres Mafla
AU  - Marçal Rusiñol
AU  - Ernest Valveny
AU  - Dimosthenis Karatzas
PY  - 2021//
TI  - Multimodal grid features and cell pointers for scene text visual question answering
T2  - PRL
JO  - Pattern Recognition Letters
SP  - 242
EP  - 249
VL  - 150
N2  - This paper presents a new model for the task of scene text visual question answering. In this task questions about a given image can only be answered by reading and understanding scene text. Current state of the art models for this task make use of a dual attention mechanism in which one attention module attends to visual features while the other attends to textual features. A possible issue with this is that it makes difficult for the model to reason jointly about both modalities. To fix this problem we propose a new model that is based on an single attention mechanism that attends to multi-modal features conditioned to the question. The output weights of this attention module over a grid of multi-modal spatial features are interpreted as the probability that a certain spatial location of the image contains the answer text to the given question. Our experiments demonstrate competitive performance in two standard datasets with a model that is  faster than previous methods at inference time. Furthermore, we also provide a novel analysis of the ST-VQA dataset based on a human performance study. Supplementary material, code, and data is made available through this link.
UR  - https://www.sciencedirect.com/science/article/pii/S0167865521002336?via%3Dihub
L1  - http://refbase.cvc.uab.es/files/GBT2021.pdf
N1  - DAG; 600.084; 600.121
ID  - Lluis Gomez2021
ER  -