AKRMap: Adaptive Kernel Regression for Trustworthy Visualization of Cross-Modal Embeddings

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ye, Yilin, Huang, Junchao, Zeng, Xingchen, Xia, Jiazhi, Zeng, Wei
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915309461438464
author Ye, Yilin
Huang, Junchao
Zeng, Xingchen
Xia, Jiazhi
Zeng, Wei
author_facet Ye, Yilin
Huang, Junchao
Zeng, Xingchen
Xia, Jiazhi
Zeng, Wei
contents Cross-modal embeddings form the foundation for multi-modal models. However, visualization methods for interpreting cross-modal embeddings have been primarily confined to traditional dimensionality reduction (DR) techniques like PCA and t-SNE. These DR methods primarily focus on feature distributions within a single modality, whilst failing to incorporate metrics (e.g., CLIPScore) across multiple modalities. This paper introduces AKRMap, a new DR technique designed to visualize cross-modal embeddings metric with enhanced accuracy by learning kernel regression of the metric landscape in the projection space. Specifically, AKRMap constructs a supervised projection network guided by a post-projection kernel regression loss, and employs adaptive generalized kernels that can be jointly optimized with the projection. This approach enables AKRMap to efficiently generate visualizations that capture complex metric distributions, while also supporting interactive features such as zoom and overlay for deeper exploration. Quantitative experiments demonstrate that AKRMap outperforms existing DR methods in generating more accurate and trustworthy visualizations. We further showcase the effectiveness of AKRMap in visualizing and comparing cross-modal embeddings for text-to-image models. Code and demo are available at https://github.com/yilinye/AKRMap.
format Preprint
id arxiv_https___arxiv_org_abs_2505_14664
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle AKRMap: Adaptive Kernel Regression for Trustworthy Visualization of Cross-Modal Embeddings
Ye, Yilin
Huang, Junchao
Zeng, Xingchen
Xia, Jiazhi
Zeng, Wei
Computer Vision and Pattern Recognition
Artificial Intelligence
Human-Computer Interaction
Machine Learning
Cross-modal embeddings form the foundation for multi-modal models. However, visualization methods for interpreting cross-modal embeddings have been primarily confined to traditional dimensionality reduction (DR) techniques like PCA and t-SNE. These DR methods primarily focus on feature distributions within a single modality, whilst failing to incorporate metrics (e.g., CLIPScore) across multiple modalities. This paper introduces AKRMap, a new DR technique designed to visualize cross-modal embeddings metric with enhanced accuracy by learning kernel regression of the metric landscape in the projection space. Specifically, AKRMap constructs a supervised projection network guided by a post-projection kernel regression loss, and employs adaptive generalized kernels that can be jointly optimized with the projection. This approach enables AKRMap to efficiently generate visualizations that capture complex metric distributions, while also supporting interactive features such as zoom and overlay for deeper exploration. Quantitative experiments demonstrate that AKRMap outperforms existing DR methods in generating more accurate and trustworthy visualizations. We further showcase the effectiveness of AKRMap in visualizing and comparing cross-modal embeddings for text-to-image models. Code and demo are available at https://github.com/yilinye/AKRMap.
title AKRMap: Adaptive Kernel Regression for Trustworthy Visualization of Cross-Modal Embeddings
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Human-Computer Interaction
Machine Learning
url https://arxiv.org/abs/2505.14664