Hystar: Hypernetwork-driven Style-adaptive Retrieval via Dynamic SVD Modulation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cai, Yujia, Li, Boxuan, Xu, Chenghao, Yan, Jiexi
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914551920852992
author Cai, Yujia
Li, Boxuan
Xu, Chenghao
Yan, Jiexi
author_facet Cai, Yujia
Li, Boxuan
Xu, Chenghao
Yan, Jiexi
contents Query-based image retrieval (QBIR) requires retrieving relevant images given diverse and often stylistically heterogeneous queries, such as sketches, artworks, or low-resolution previews. While large-scale vision--language representation models (VLRMs) like CLIP offer strong zero-shot retrieval performance, they struggle with distribution shifts caused by unseen query styles. In this paper, we propose the Hypernetwork-driven Style-adaptive Retrieval (Hystar), a lightweight framework that dynamically adapts model weights to each query's style. Hystar employs a hypernetwork to generate singular-value perturbations ($ΔS$) for attention layers, enabling flexible per-input adaptation, while static singular-value offsets on MLP layers ensure cross-style stability. To better handle semantic confusions across styles, we design StyleNCE as part of Hystar, an optimal-transport-weighted contrastive loss that emphasizes hard cross-style negatives. Extensive experiments on multi-style retrieval and cross-style classification benchmarks demonstrate that Hystar consistently outperforms strong baselines, achieving state-of-the-art performance while being parameter-efficient and stable across styles.
format Preprint
id arxiv_https___arxiv_org_abs_2605_10009
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Hystar: Hypernetwork-driven Style-adaptive Retrieval via Dynamic SVD Modulation
Cai, Yujia
Li, Boxuan
Xu, Chenghao
Yan, Jiexi
Computer Vision and Pattern Recognition
Query-based image retrieval (QBIR) requires retrieving relevant images given diverse and often stylistically heterogeneous queries, such as sketches, artworks, or low-resolution previews. While large-scale vision--language representation models (VLRMs) like CLIP offer strong zero-shot retrieval performance, they struggle with distribution shifts caused by unseen query styles. In this paper, we propose the Hypernetwork-driven Style-adaptive Retrieval (Hystar), a lightweight framework that dynamically adapts model weights to each query's style. Hystar employs a hypernetwork to generate singular-value perturbations ($ΔS$) for attention layers, enabling flexible per-input adaptation, while static singular-value offsets on MLP layers ensure cross-style stability. To better handle semantic confusions across styles, we design StyleNCE as part of Hystar, an optimal-transport-weighted contrastive loss that emphasizes hard cross-style negatives. Extensive experiments on multi-style retrieval and cross-style classification benchmarks demonstrate that Hystar consistently outperforms strong baselines, achieving state-of-the-art performance while being parameter-efficient and stable across styles.
title Hystar: Hypernetwork-driven Style-adaptive Retrieval via Dynamic SVD Modulation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2605.10009