Research map / Vision-language and multimodal / CLIP and zero-shot
CLAP: Isolating Content from Style through Contrastive Learning with Augmented Prompts
Yichao Cai, Yuhang Liu, Zhen Zhang, Javen Qinfeng Shi
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 contrastive learning with augmented prompts (CLAP) to CLIP representations | CLIP ViT-B/16 · Zero-shot ZS(C) overall | avg top-1 accuracy (%): 77.4 -> 79.9 | Improved |
| apply CLAP to CLIP under noised prompt | CLIP ViT-B/16 · Zero-shot ZS(NC) overall | avg top-1 accuracy (%): 70.4 -> 76.1 | Improved |
| use image augmentation (Im.Aug) instead of prompt augmentation (ablation) | CLIP ViT-B/16 · Zero-shot ZS(C) overall | avg top-1 accuracy (%): 77.4 -> 76.1 | Hurt |
| use image augmentation (Im.Aug) on OfficeHome (ablation) | CLIP ViT-B/16 · OfficeHome ZS(C) | top-1 accuracy (%): 79.8 -> 77.0 | Mixed |
| switch from image augmentation to prompt augmentation (ablation) | CLAP · Prompt variance overall | range R: 1.9 -> 1.1 | Improved |
| apply CLAP | CLIP ViT-B/16 · Few-shot 1-shot linear probe PACS | top-1 accuracy gain: +10% | Improved |
| apply CLAP to 1-shot classifier under PGD-20 | CLIP ViT-B/16 · Adversarial robustness 1-shot PGD-20 | accuracy gain: +11.7% | Improved |
| apply CLAP to zero-shot classifier under CW-20 | CLIP ViT-B/16 · Adversarial robustness ZS(C) CW-20 | accuracy gain: +1.1% | No clear effect |
| apply Im.Aug to zero-shot classifier under CW-20 (ablation) | CLIP ViT-B/16 · Adversarial robustness ZS(C) CW-20 | accuracy gain: 0% | No clear effect |
| vary latent dimension, temperature and inference weight (ablation) | CLAP · VLCS zero-shot | avg top-1 accuracy | No clear effect |
| use simpler training prompt forms rather than LLM or random-style prompts (ablation) | CLAP · VLCS zero-shot | avg top-1 accuracy | Improved |
| train disentangled network on text prompts instead of synthetic images (ablation) | CLAP · Training time DomainNet | training time: 3.3 hours -> about 47 minutes | Improved |
Later papers that compared with this method
11 later accepted papers report CLAP as a baseline in their results tables; 7 of them report a better number on most of the results they share with it.
- BayesAdapter: Enhanced Uncertainty Estimation in CLIP Few-Shot Adaptation (IJCV 2026 · better on 6 of 8 shared results)
- ORION: ORthonormal Text Encoding for Universal VLM AdaptatION (CVPR 2026 · better on 3 of 3 shared results)
- Efficient and High-Fidelity Omni Modality Retrieval (CVPR 2026 · better on 2 of 2 shared results)
- Meric: A Unified Framework for Multimodal Music Generation and Retrieval via Representation Space Anchoring (ECCV 2026 · better on 1 of 2 shared results)
- SAM Audio: Segment Anything in Audio (ICML 2026 · better on 1 of 1 shared result)
- Subspace Alignment for CLIP-based Continual Learning via Canonical Correlation Analysis (CVPR 2026 · better on 1 of 1 shared result)
- Text and Image Are Mutually Beneficial: Enhancing Training-Free Few-Shot Classification with CLIP (AAAI 2025 · better on 0 of 2 shared results)
- ProKeR: A Kernel Perspective on Few-Shot Adaptation of Large Vision-Language Models (CVPR 2025 · better on 1 of 1 shared result)
3 more not listed.
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Retrieval-Enhanced Contrastive Vision-Text Models (ICLR 2024 · similarity 0.89)
- Diffusion Feedback Helps CLIP See Better (ICLR 2025 · similarity 0.88)
- Understanding the Robustness of Multi-modal Contrastive Learning to Distribution Shift (ICLR 2024 · similarity 0.88)
- CLIPLoss and Norm-Based Data Selection Methods for Multimodal Contrastive Learning (NeurIPS 2024 · similarity 0.88)
- Progressive Compositionality in Text-to-Image Generative Models (ICLR 2025 · similarity 0.87)
- Modeling Caption Diversity in Contrastive Vision-Language Pretraining (ICML 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.
