Lune

EMNLP Findings2021

Retrieval, Analogy, and Composition: A framework for Compositional Generalization in Image Captioning

Zhan Shi, Hui Liu, Martin Renqiang Min, Christopher Malon, Li Erran Li, Xiaodan Zhu

Abstract

Image captioning systems are expected to have the ability to combine individual concepts when describing scenes with concept combinations that are not observed during training. In spite of significant progress in image captioning with the help of the autoregressive generation framework, current approaches fail to generalize well to novel concept combinations. We propose a new framework that revolves around probing several similar image caption training instances (retrieval), performing analogical reasoning over relevant entities in retrieved prototypes (analogy), and enhancing the generation process with reasoning outcomes (composition). Our method augments the generation model by referring to the neighboring instances in the training set to produce novel concept combinations in generated captions. We perform experiments on the widely used image captioning benchmarks. The proposed models achieve substantial improvement over the compared baselines on both composition related evaluation metrics and conventional image captioning metrics.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

1 top-tier paper has cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers1

  1. Compositional Causal Reasoning Evaluation in Language ModelsJacqueline R. M. A. Maasch, Alihan Hüyük, Xinnuo Xu, Aditya V. Nori et al.ICML 2025

Builds on8

  1. Learning Transferable Visual Models From Natural Language SupervisionAlec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh et al.ICML 2021 · 47,906 citations
  2. The Curious Case of Neural Text DegenerationAri Holtzman, Jan Buys, Li Du, Maxwell Forbes et al.ICLR 2020 · 4,112 citations
  3. Attention on Attention for Image CaptioningLun Huang, Wenmin Wang, Jie Chen, Xiaoyong WeiICCV 2019 · 992 citations
  4. Measuring Compositional Generalization: A Comprehensive Method on Realistic DataDaniel Keysers, Nathanael Schärli, Nathan Scales, Hylke Buisman et al.ICLR 2020 · 401 citations
  5. Improving Image Captioning with Better Use of CaptionZhan Shi, Xu Zhou, Xipeng Qiu, Xiaodan ZhuACL 2020 · 84 citations

Related papers

  1. Interactive Dual Generative Adversarial Networks for Image CaptioningJunhao Liu, Kai Wang, Chunpu Xu, Zhou Zhao et al.AAAI 2020 · 35 citations
  2. Set Prediction Guided by Semantic Concepts for Diverse Video CaptioningYifan Lu, Ziqi Zhang, Chunfeng Yuan, Peng Li et al.AAAI 2024 · 7 citations
  3. Generative Thinking, Corrective Action: User-Friendly Composed Image Retrieval via Automatic Multi-Agent CollaborationZhangtao Cheng, Yuhao Ma, Jian Lang, Kunpeng Zhang et al.KDD 2025 · 2 citations
  4. Learning by Analogy: A Causal Framework for Compositional GeneralizationLingjing Kong, Shaoan Xie, Yang Jiao, Yetian Chen et al.CVPR 2026
  5. Towards Accurate Text-Based Image Captioning With Content Diversity ExplorationGuanghui Xu, Shuaicheng Niu, Mingkui Tan, Yucheng Luo et al.CVPR 2021
Dusk over the sea between two cliffs drawn in fine vertical lines