AI Paper Atlas

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

MMLoP: Multi-Modal Low-Rank Prompting for Efficient Vision-Language Adaptation

Sajjad Ghiasvand, Haniyeh Oskouie, Mahnoosh Alizadeh, Ramtin Pedarsani

ECCV 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Prompt learning for CLIP · 6 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
add rank-1 low-rank prompt factorization (ablation)IVLP baseline · Base-to-novel (11 datasets average)harmonic mean: 77.51 -> 77.06Hurt
add self-regulating consistency loss (SCL) (ablation)low-rank IVLP · Base-to-novel (11 datasets average)novel accuracy: 71.39 -> 74.48Improved
add self-regulating consistency loss (SCL) (ablation)low-rank IVLP · Base-to-novel (11 datasets average)harmonic mean: 77.06 -> 78.78Improved
add uniform drift correction (UDC) (ablation)low-rank IVLP + SCL · Base-to-novel (11 datasets average)harmonic mean: 78.78 -> 79.20Improved
add shared up-projection (ablation)low-rank IVLP + SCL + UDC · Base-to-novel (11 datasets average)harmonic mean: 79.20 -> 79.70Improved
full MMLoP vs IVLPIVLP · Base-to-novel (11 datasets average)base accuracy: 84.21 -> 83.79Hurt
full MMLoP vs IVLPIVLP · Base-to-novel (11 datasets average)novel accuracy: 71.79 -> 75.98Improved
prompt depth 7 instead of 9 (ablation)MMLoP · Base-to-novel (9 datasets average)harmonic mean: 79.91 -> 78.63Hurt
prompt depth 11 instead of 9 (ablation)MMLoP · Base-to-novel (9 datasets average)novel accuracy: 75.85 -> 74.58Hurt
prompt length 2 or 8 instead of 4 (ablation)MMLoP · Base-to-novel (9 datasets average)harmonic mean: 79.91 -> 79.31 / 79.26No clear effect
rank 2 or 4 instead of 1 (ablation)MMLoP · Base-to-novel (9 datasets average)novel accuracy: 75.85 -> 75.01 / 75.20Hurt
compare MMLoP with 51x more parameters method MMAMMA · Domain generalizationaverage accuracy: 60.46 vs 60.48No clear effect

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.

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