Research map / Vision-language and multimodal / CLIP and zero-shot
SDPT: Synchronous Dual Prompt Tuning for Visual-Language Pre-Trained Models
Yang Zhou, Yongjian Wu, Jiya Saiyin, Bingzheng Wei, Yan Xu
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 |
|---|---|---|---|
| use shared unified prototype tokens instead of unshared tokens (ablation) | SDPT · Aquarium | mAP: 53.3 -> 57.7 | Improved |
| train both modalities synchronously instead of asynchronously (ablation) | SDPT · Aquarium | mAP: 54.8 -> 57.7 | Improved |
| use inverse linear projections instead of learnable projections (dual) (ablation) | SDPT · Aquarium | mAP: 54.3 -> 57.7 | Improved |
| apply dual-modal instead of image-only inverse projection (ablation) | SDPT · Aquarium | mAP: 53.1 -> 57.7 | Improved |
| apply dual-modal instead of text-only inverse projection (ablation) | SDPT · Aquarium | mAP: 55.4 -> 57.7 | Improved |
| vary token length k from 10 to 200 (ablation) | SDPT · Aquarium | mAP: 57.1 -> 57.7 | No clear effect |
| insert tokens in all 8 layers instead of layers 1-4 (ablation) | SDPT · Aquarium | mAP: 57.1 -> 57.7 | Improved |
| insert tokens in only layer 8 (ablation) | SDPT · Aquarium | mAP: 57.7 -> 57.5 | No clear effect |
| tune X-MHA with LoRA instead of encoders | LoRA · COCO | mAP: 56.5 -> 55.4 | Hurt |
| tune X-MHA with BitFit instead of encoders | BitFit · COCO | mAP: 54.5 -> 52.6 | Hurt |
| SDPT k=10 versus full fine-tuning | GLIP-L · LVIS | mAP: 41.2 -> 41.2 | No clear effect |
| SDPT instead of new Adapter on frozen COCO adapter | GLIP-L · COCO old-task mAP | mAP: 35.4 -> 53.2 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MaPLe — MaPLe: Multi-Modal Prompt Learning (CVPR 2023) · better on 4 of 4 shared results
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
- Amend to Alignment: Decoupled Prompt Tuning for Mitigating Spurious Correlation in Vision-Language Models (ICML 2024 · similarity 0.89)
- SA²VP: Spatially Aligned-and-Adapted Visual Prompt (AAAI 2024 · similarity 0.88)
- DPC: Dual-Prompt Collaboration for Tuning Vision-Language Models (CVPR 2025 · similarity 0.88)
- COMMA: Co-articulated Multi-Modal Learning (AAAI 2024 · similarity 0.88)
- MmAP: Multi-Modal Alignment Prompt for Cross-Domain Multi-Task Learning (AAAI 2024 · similarity 0.88)
- Consistency-guided Prompt Learning for Vision-Language Models (ICLR 2024 · similarity 0.88)
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.
