Research map / Vision-language and multimodal / Human and embodied modeling
Unsupervised Image Captioning
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add the initialization pipeline (ablation) | Unsupervised image captioning model · MSCOCO test split | CIDEr: 22.9 -> 28.6 | Helped |
| Add the initialization pipeline (ablation) | Unsupervised image captioning model · MSCOCO test split, unpaired setting | CIDEr: 46.7 -> 54.9 | Helped |
| Add adversarial training to the feat2sen initialization (ablation) | Unpaired image captioning model · MSCOCO test split, unpaired setting | CIDEr: 35.4 -> 45.8 | Helped |
| Add concept reward to adversarial training (ablation) | Unpaired image captioning model · MSCOCO test split, unpaired setting | CIDEr: 45.8 -> 48.8 | Helped |
| Add image reconstruction objective to adversarial training and concept reward (ablation) | Unpaired image captioning model · MSCOCO test split, unpaired setting | CIDEr: 48.8 -> 49.0 | Helped |
| Add sentence reconstruction to adversarial training, concept reward, and image reconstruction (ablation) | Unpaired image captioning model · MSCOCO test split, unpaired setting | CIDEr: 49 -> 54.9 | Helped |
| Use adversarial training alone rather than the full method (ablation) | Unsupervised image captioning model · MSCOCO test split | CIDEr: 28.6 -> 11.8 | Hurt |
| Use adversarial training alone rather than the full method (ablation) | Unpaired image captioning model · MSCOCO test split, unpaired setting | CIDEr: 54.9 -> 45.8 | Hurt |
Closest papers by meaning
- Partially-Supervised Image Captioning (NeurIPS 2018 · similarity 0.90)
- Deep Compositional Captioning: Describing Novel Object Categories Without Paired Training Data (CVPR 2016 · similarity 0.90)
- Convolutional Image Captioning (CVPR 2018 · similarity 0.89)
- Relational Distant Supervision for Image Captioning without Image-Text Pairs (AAAI 2024 · similarity 0.89)
- Beyond a Pre-Trained Object Detector: Cross-Modal Textual and Visual Context for Image Captioning (CVPR 2022 · similarity 0.89)
- Captioning Images With Diverse Objects (CVPR 2017 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.