Federated Cross-Modal Style-Aware Prompt Generation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Prasad, Suraj, Mahla, Navyansh, Gupta, Sunny, Sethi, Amit
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911109235081216
author Prasad, Suraj
Mahla, Navyansh
Gupta, Sunny
Sethi, Amit
author_facet Prasad, Suraj
Mahla, Navyansh
Gupta, Sunny
Sethi, Amit
contents Prompt learning has propelled vision-language models like CLIP to excel in diverse tasks, making them ideal for federated learning due to computational efficiency. However, conventional approaches that rely solely on final-layer features miss out on rich multi-scale visual cues and domain-specific style variations in decentralized client data. To bridge this gap, we introduce FedCSAP (Federated Cross-Modal Style-Aware Prompt Generation). Our framework harnesses low, mid, and high-level features from CLIP's vision encoder alongside client-specific style indicators derived from batch-level statistics. By merging intricate visual details with textual context, FedCSAP produces robust, context-aware prompt tokens that are both distinct and non-redundant, thereby boosting generalization across seen and unseen classes. Operating within a federated learning paradigm, our approach ensures data privacy through local training and global aggregation, adeptly handling non-IID class distributions and diverse domain-specific styles. Comprehensive experiments on multiple image classification datasets confirm that FedCSAP outperforms existing federated prompt learning methods in both accuracy and overall generalization.
format Preprint
id arxiv_https___arxiv_org_abs_2508_12399
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Federated Cross-Modal Style-Aware Prompt Generation
Prasad, Suraj
Mahla, Navyansh
Gupta, Sunny
Sethi, Amit
Computer Vision and Pattern Recognition
Prompt learning has propelled vision-language models like CLIP to excel in diverse tasks, making them ideal for federated learning due to computational efficiency. However, conventional approaches that rely solely on final-layer features miss out on rich multi-scale visual cues and domain-specific style variations in decentralized client data. To bridge this gap, we introduce FedCSAP (Federated Cross-Modal Style-Aware Prompt Generation). Our framework harnesses low, mid, and high-level features from CLIP's vision encoder alongside client-specific style indicators derived from batch-level statistics. By merging intricate visual details with textual context, FedCSAP produces robust, context-aware prompt tokens that are both distinct and non-redundant, thereby boosting generalization across seen and unseen classes. Operating within a federated learning paradigm, our approach ensures data privacy through local training and global aggregation, adeptly handling non-IID class distributions and diverse domain-specific styles. Comprehensive experiments on multiple image classification datasets confirm that FedCSAP outperforms existing federated prompt learning methods in both accuracy and overall generalization.
title Federated Cross-Modal Style-Aware Prompt Generation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.12399