AI Paper Atlas

Research map / Federated learning and privacy / Personalized and heterogeneous FL

Federated Prompt-Tuning with Heterogeneous and Incomplete Multimodal Client Data

Thu Hang Phung, Duong M. Nguyen, Thanh Trung Huynh, Quoc Viet Hung Nguyen, Trong Nghia Hoang, Phi Le Nguyen

ICCV 2025 · Federated learning and privacy › Personalized and heterogeneous FL · approach: Federated prompt and LoRA tuning · 1 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 inter-client prompt set (FED-PRIME vs FED-INTRA only) (ablation)FED-PRIME · UPMC Food-101 train Miss Text, test ~Trainaccuracy: 61.71 -> 78.88Improved
add intra-client prompt set (FED-PRIME vs FED-INTER only) (ablation)FED-PRIME · UPMC Food-101 train Miss Text, test ~Trainaccuracy: 54.82 -> 78.88Improved
add inter-client prompt set (ablation)FED-PRIME · UPMC Food-101 train Miss Image, test ~Trainaccuracy: 22.84 -> 90.55Improved
add intra-client prompt set (ablation)FED-PRIME · MM-IMDB train Miss Both, test ~TrainF1-macro: 26.53 -> 32.01Improved
add inter-client prompt set (ablation)FED-PRIME · MM-IMDB train Miss Both, test ~TrainF1-macro: 11.90 -> 32.01Improved
increase prompt pool size from 10 to 20 (ablation)FED-PRIME · UPMC Food-101top-1 accuracy: 73.94 -> 82.61Improved
increase prompt pool size from 20 to 30 (ablation)FED-PRIME · UPMC Food-101top-1 accuracy: 82.61 -> 83.21No clear effect
use FED-PRIME instead of best baseline FedAvg-style prompt tuningFEDAVG-P · UPMC Food-101 train Miss Both, test ~Trainaccuracy: 56.32 -> 84.44Improved
use FED-PRIME instead of FEDMSPLIT-P on Miss Image trainingFEDMSPLIT-P · UPMC Food-101 train Miss Image, test ~Trainaccuracy: 74.16 -> 90.55Improved
prepend prompts only to first MSA layer instead of first six layersFEDAVG-P · UPMC Food-101GPU memory (GB): 31.70 -> 16.80Improved
test with image only after Miss Image trainingFED-PRIME · UPMC Food-101accuracy: 90.55 -> 54.14Hurt

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