Research map / Vision-language and multimodal / Human and embodied modeling
Contrastive Learning for Image Captioning
Bo Dai, Dahua Lin
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Apply CL to AdaptiveAttention | AdaptiveAttention · COCO online testing server C40 | CIDEr: 1.004 -> 1.029 | Improved |
| Apply CL to AdaptiveAttention | AdaptiveAttention · InstaPIC-1.1M test | CIDEr: 0.126 -> 0.144 | Improved |
| Use only the positive constraint instead of full CL (ablation) | AdaptiveAttention + CL · MSCOCO validation | CIDEr: 1.142 -> 1.059 | Hurt |
| Use only the negative constraint instead of full CL (ablation) | AdaptiveAttention + CL · MSCOCO validation | CIDEr: 1.142 -> 0.603 | Hurt |
| Add the positive constraint to negative-constraint-only CL (ablation) | AdaptiveAttention + CL(N) · MSCOCO validation | CIDEr: 0.603 -> 1.142 | Improved |
| Train with IL instead of MLE | AdaptiveAttention · MSCOCO validation | CIDEr: 1.042 -> 1.004 | Hurt |
| Train with GAN instead of MLE | AdaptiveAttention · MSCOCO validation | CIDEr: 1.042 -> 0.700 | Hurt |
| Use AdaptiveAttention rather than Neuraltalk2 as the reference | Neuraltalk2 + CL · MSCOCO validation | CIDEr: 0.905 -> 0.956 | Improved |
| Replace the reference with the best target from run 1 (ablation) | AdaptiveAttention + CL · MSCOCO validation | CIDEr: 1.142 -> 1.142 | No clear effect |
Closest papers by meaning
- On Distinctive Image Captioning via Comparing and Reweighting (TPAMI 2022 · similarity 0.89)
- Modeling Caption Diversity in Contrastive Vision-Language Pretraining (ICML 2024 · similarity 0.89)
- FFF: Fixing Flawed Foundations in Contrastive Pre-Training Results in Very Strong Vision-Language Models (CVPR 2024 · similarity 0.88)
- SyCoCa: Symmetrizing Contrastive Captioners with Attentive Masking for Multimodal Alignment (ICML 2024 · similarity 0.88)
- Mining Fine-Grained Image-Text Alignment for Zero-Shot Captioning via Text-Only Training (AAAI 2024 · similarity 0.87)
- Noise-Aware Image Captioning with Progressively Exploring Mismatched Words (AAAI 2024 · similarity 0.87)
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.
