Geometry Attention Transformer with Position-aware LSTMs for Image Captioning [article]

Chi Wang, Yulin Shen, Luping Ji
2021 arXiv   pre-print
In recent years, transformer structures have been widely applied in image captioning with impressive performance. For good captioning results, the geometry and position relations of different visual objects are often thought of as crucial information. Aiming to further promote image captioning by transformers, this paper proposes an improved Geometry Attention Transformer (GAT) model. In order to further leverage geometric information, two novel geometry-aware architectures are designed
more » ... vely for the encoder and decoder in our GAT. Besides, this model includes the two work modules: 1) a geometry gate-controlled self-attention refiner, for explicitly incorporating relative spatial information into image region representations in encoding steps, and 2) a group of position-LSTMs, for precisely informing the decoder of relative word position in generating caption texts. The experiment comparisons on the datasets MS COCO and Flickr30K show that our GAT is efficient, and it could often outperform current state-of-the-art image captioning models.
arXiv:2110.00335v1 fatcat:emucqxpc3rdfpeu3xoewpwuj2i