Proceedings

Leveraging Human Attention in Novel Object Captioning

Xianyu Chen, Ming Jiang, Qi Zhao

Download BibTeX Cite

Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence. Main Track. Pages 622-628.

https://doi.org/10.24963/ijcai.2021/86

Keywords

  • Computer Vision: Language and Vision

Abstract

Image captioning models depend on training with paired image-text corpora, which poses various challenges in describing images containing novel objects absent from the training data. While previous novel object captioning methods rely on external image taggers or object detectors to describe novel objects, we present the Attention-based Novel Object Captioner (ANOC) that complements novel object captioners with human attention features that characterize generally important information independent of tasks. It introduces a gating mechanism that adaptively incorporates human attention with self-learned machine attention, with a Constrained Self-Critical Sequence Training method to address the exposure bias while maintaining constraints of novel object descriptions. Extensive experiments conducted on the nocaps and Held-Out COCO datasets demonstrate that our method considerably outperforms the state-of-the-art novel object captioners.

Our source code is available at https://github.com/chenxy99/ANOC.

Xianyu Chen, Ming Jiang, and Qi Zhao. 2021. Leveraging Human Attention in Novel Object Captioning. In Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence, IJCAI-21. IJCAI, 622–628. https://doi.org/10.24963/ijcai.2021/86
Welcome Future Conferences AI Journal Division IJCAI Proceedings About IJCAI IJCAI Awards