Cross-Modal Similarity-Based Curriculum Learning for Image Captioning
Hongkuan Zhang, Saku Sugawara, Akiko Aizawa, Lei Zhou, Ryohei Sasano, Koichi Takeda
Abstract
Image captioning models require the high-level generalization ability to describe the contents of various images in words. Most existing approaches treat the image–caption pairs equally in their training without considering the differences in their learning difficulties. Several image captioning approaches introduce curriculum learning methods that present training data with increasing levels of difficulty. However, their difficulty measurements are either based on domain-specific features or prior model training. In this paper, we propose a simple yet efficient difficulty measurement for image captioning using cross-modal similarity calculated by a pretrained vision–language model. Experiments on the COCO and Flickr30k datasets show that our proposed approach achieves superior performance and competitive convergence speed to baselines without requiring heuristics or incurring additional training costs. Moreover, the higher model performance on difficult examples and unseen data also demonstrates the generalization ability.- Anthology ID:
- 2022.emnlp-main.516
- Volume:
- Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
- Month:
- December
- Year:
- 2022
- Address:
- Abu Dhabi, United Arab Emirates
- Editors:
- Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
- Venue:
- EMNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 7599–7606
- Language:
- URL:
- https://aclanthology.org/2022.emnlp-main.516
- DOI:
- 10.18653/v1/2022.emnlp-main.516
- Cite (ACL):
- Hongkuan Zhang, Saku Sugawara, Akiko Aizawa, Lei Zhou, Ryohei Sasano, and Koichi Takeda. 2022. Cross-Modal Similarity-Based Curriculum Learning for Image Captioning. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 7599–7606, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
- Cite (Informal):
- Cross-Modal Similarity-Based Curriculum Learning for Image Captioning (Zhang et al., EMNLP 2022)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-1/2022.emnlp-main.516.pdf