Abstract
In this paper, we study the problem of recognizing compositional attribute-object concepts within the zero-shot learning (ZSL) framework. We propose an episode-based cross-attention (EpiCA) network which combines merits of cross-attention mechanism and episode-based training strategy to recognize novel compositional concepts. Firstly, EpiCA bases on cross-attention to correlate conceptvisual information and utilizes the gated pooling layer to build contextualized representations for both images and concepts. The updated representations are used for a more indepth multi-modal relevance calculation for concept recognition. Secondly, a two-phase episode training strategy, especially the ransductive phase, is adopted to utilize unlabeled test examples to alleviate the low-resource learning problem. Experiments on two widelyused zero-shot compositional learning (ZSCL) benchmarks have demonstrated the effectiveness of the model compared with recent approaches on both conventional and generalized ZSCL settings.- Anthology ID:
- 2021.metanlp-1.3
- Volume:
- Proceedings of the 1st Workshop on Meta Learning and Its Applications to Natural Language Processing
- Month:
- August
- Year:
- 2021
- Address:
- Online
- Editors:
- Hung-Yi Lee, Mitra Mohtarami, Shang-Wen Li, Di Jin, Mandy Korpusik, Shuyan Dong, Ngoc Thang Vu, Dilek Hakkani-Tur
- Venue:
- MetaNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 19–27
- Language:
- URL:
- https://aclanthology.org/2021.metanlp-1.3
- DOI:
- 10.18653/v1/2021.metanlp-1.3
- Cite (ACL):
- Guangyue Xu, Parisa Kordjamshidi, and Joyce Chai. 2021. Zero-Shot Compositional Concept Learning. In Proceedings of the 1st Workshop on Meta Learning and Its Applications to Natural Language Processing, pages 19–27, Online. Association for Computational Linguistics.
- Cite (Informal):
- Zero-Shot Compositional Concept Learning (Xu et al., MetaNLP 2021)
- PDF:
- https://preview.aclanthology.org/ingest-2024-clasp/2021.metanlp-1.3.pdf
- Data
- MIT-States