GSE: Evaluating Sticker Visual Semantic Similarity via a General Sticker Encoder

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chee, Heng Er Metilda, Wang, Jiayin, Guo, Zhiqiang, Ma, Weizhi, Zhang, Min
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915604079837184
author Chee, Heng Er Metilda
Wang, Jiayin
Guo, Zhiqiang
Ma, Weizhi
Zhang, Min
author_facet Chee, Heng Er Metilda
Wang, Jiayin
Guo, Zhiqiang
Ma, Weizhi
Zhang, Min
contents Stickers have become a popular form of visual communication, yet understanding their semantic relationships remains challenging due to their highly diverse and symbolic content. In this work, we formally {define the Sticker Semantic Similarity task} and introduce {Triple-S}, the first benchmark for this task, consisting of 905 human-annotated positive and negative sticker pairs. Through extensive evaluation, we show that existing pretrained vision and multimodal models struggle to capture nuanced sticker semantics. To address this, we propose the {General Sticker Encoder (GSE)}, a lightweight and versatile model that learns robust sticker embeddings using both Triple-S and additional datasets. GSE achieves superior performance on unseen stickers, and demonstrates strong results on downstream tasks such as emotion classification and sticker-to-sticker retrieval. By releasing both Triple-S and GSE, we provide standardized evaluation tools and robust embeddings, enabling future research in sticker understanding, retrieval, and multimodal content generation. The Triple-S benchmark and GSE have been publicly released and are available here.
format Preprint
id arxiv_https___arxiv_org_abs_2511_04977
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GSE: Evaluating Sticker Visual Semantic Similarity via a General Sticker Encoder
Chee, Heng Er Metilda
Wang, Jiayin
Guo, Zhiqiang
Ma, Weizhi
Zhang, Min
Computer Vision and Pattern Recognition
Multimedia
Stickers have become a popular form of visual communication, yet understanding their semantic relationships remains challenging due to their highly diverse and symbolic content. In this work, we formally {define the Sticker Semantic Similarity task} and introduce {Triple-S}, the first benchmark for this task, consisting of 905 human-annotated positive and negative sticker pairs. Through extensive evaluation, we show that existing pretrained vision and multimodal models struggle to capture nuanced sticker semantics. To address this, we propose the {General Sticker Encoder (GSE)}, a lightweight and versatile model that learns robust sticker embeddings using both Triple-S and additional datasets. GSE achieves superior performance on unseen stickers, and demonstrates strong results on downstream tasks such as emotion classification and sticker-to-sticker retrieval. By releasing both Triple-S and GSE, we provide standardized evaluation tools and robust embeddings, enabling future research in sticker understanding, retrieval, and multimodal content generation. The Triple-S benchmark and GSE have been publicly released and are available here.
title GSE: Evaluating Sticker Visual Semantic Similarity via a General Sticker Encoder
topic Computer Vision and Pattern Recognition
Multimedia
url https://arxiv.org/abs/2511.04977