AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

Soft Prompt Generation for Domain Generalization

Shuanghao Bai, Yuedi Zhang, Wanqi Zhou, Zhirong Luan, Badong Chen

ECCV 2024 · Vision-language and multimodal › CLIP and zero-shot · approach: Prompt learning for CLIP · 30 citations (OpenAlex)

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 changedWhereMeasuredEffect
replace fixed/conditional prompts with prompts generated by a CGAN (ablation)CLIP RN50 prompt learning · Multi-source DG average of 5 datasetsaccuracy: 66.3 -> 67.5 (vs CoCoOp)Improved
use conditional prompts (CoCoOp, DPL) instead of a single fixed learned prompt (ablation)CLIP RN50 prompt learning · Multi-source DG, TerraInc.accuracy: 33.2 -> 34.1Improved
learn prompts instead of a manual prompt (ablation)CLIP RN50 · TerraIncognita multi-source DGaccuracy: 23.8 -> 33.2Improved
average per-domain prompts (mix-domain) instead of one prompt from all domains (ablation)fixed learned prompt · VLCS multi-source DGaccuracy: 81.4 -> 76.2Hurt
set prompt context length to 4 rather than 2, 8 or 16 (ablation)SPG · PACS multi-source DGaverage accuracy: 8: 88.3 -> 4: 92.8Improved
random Gaussian init instead of word-embedding init of the prompt (ablation)SPG · PACS multi-source DGaverage accuracy: 92.8 -> 92.2No clear effect
train the CGAN on more of the training data (ablation)SPG · PACS multi-source DGaverage accuracy: 30.7 (20%) -> 92.8 (full)Improved
use SPG vs traditional DG methods on TerraIncognitaCLIP RN50 vs ResNet50 DG methods · TerraIncognita multi-source DGaccuracy: SPG 37.5 vs MIRO 50.4Hurt
use SPG in cross-dataset generalization (train on DomainNet)CLIP ViT-B/16 · VLCS cross-datasetaccuracy: 80.5 (ZS-CLIP) -> 76.7Hurt
use SPG on cross-dataset generalization averageCLIP ViT-B/16 · Cross-dataset, 4 datasets averageaccuracy: 72.0 -> 73.8Improved
use SPG in single-source DGCLIP RN50 · Single-source DG, 5-dataset averageaccuracy: 61.4 -> 62.8 (vs CoOp)Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most 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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy