Learning from Children: Improving Image-Caption Pretraining via Curriculum

Hammad Ayyubi, Rahul Lokesh, Alireza Zareian, Bo Wu, Shih-Fu Chang


Abstract
Image-caption pretraining has been quite successfully used for downstream vision tasks like zero-shot image classification and object detection. However, image-caption pretraining is still a hard problem – it requires multiple concepts (nouns) from captions to be aligned to several objects in images. To tackle this problem, we go to the roots – the best learner, children. We take inspiration from cognitive science studies dealing with children’s language learning to propose a curriculum learning framework. The learning begins with easy-to-align image caption pairs containing one concept per caption. The difficulty is progressively increased with each new phase by adding one more concept per caption. Correspondingly, the knowledge acquired in each learning phase is utilized in subsequent phases to effectively constrain the learning problem to aligning one new concept-object pair in each phase. We show that this learning strategy improves over vanilla image-caption training in various settings – pretraining from scratch, using a pretrained image or/and pretrained text encoder, low data regime etc.
Anthology ID:
2023.findings-acl.846
Volume:
Findings of the Association for Computational Linguistics: ACL 2023
Month:
July
Year:
2023
Address:
Toronto, Canada
Editors:
Anna Rogers, Jordan Boyd-Graber, Naoaki Okazaki
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
13378–13386
Language:
URL:
https://aclanthology.org/2023.findings-acl.846
DOI:
10.18653/v1/2023.findings-acl.846
Bibkey:
Cite (ACL):
Hammad Ayyubi, Rahul Lokesh, Alireza Zareian, Bo Wu, and Shih-Fu Chang. 2023. Learning from Children: Improving Image-Caption Pretraining via Curriculum. In Findings of the Association for Computational Linguistics: ACL 2023, pages 13378–13386, Toronto, Canada. Association for Computational Linguistics.
Cite (Informal):
Learning from Children: Improving Image-Caption Pretraining via Curriculum (Ayyubi et al., Findings 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.findings-acl.846.pdf
Video:
 https://aclanthology.org/2023.findings-acl.846.mp4