Modeling Caption Diversity In Contrastive Vision-language Pretraining

Abstract

There are a thousand ways to caption an image. Contrastive Language Pretraining (CLIP) on the other hand, works by mapping an image and its caption to a single vector -- limiting how well CLIP-like models can represent the diverse ways to describe an image. In this work, we introduce Llip, Latent Language Image Pretraining, which models the diversity of captions that could match an image. Llip's vision encoder outputs a set of visual features that are mixed into a final representation by conditioning on information derived from the text. We show that Llip outperforms non-contextualized baselines like CLIP and SigLIP on a variety of tasks even with large-scale encoders. Llip improves zero-shot classification by an average of 2.9% zero-shot classification benchmarks with a ViT-G/14 encoder. Specifically, Llip attains a zero-shot top-1 accuracy of 83.5% on ImageNet outperforming a similarly sized CLIP by 1.4%. We also demonstrate improvement on zero-shot retrieval on MS-COCO by 6.0%. We p

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keylavoie2024modeling

Related papers

• Advancing Myopia To Holism: Fully Contrastive Language-image Pre-training (2024)0.00
• Dreamlip: Language-image Pre-training With Long Captions (2024)10.61
• Semantic Compositions Enhance Vision-language Contrastive Learning (2024)0.00
• Contrastive Vision-language Learning With Paraphrasing And Negation (2025)0.00
• Clip-lite: Information Efficient Visual Representation Learning With Language Supervision (2021)2.35
• Tripletclip: Improving Compositional Reasoning Of CLIP Via Synthetic Vision-language Negatives (2024)4.52
• LLM2CLIP: Powerful Language Model Unlocks Richer Cross-modality Representation (2024)2.26
• CLIP Is Shortsighted: Paying Attention Beyond The First Sentence (2026)0.00