A Tour of Visualization Techniques for Computer Vision Datasets

TL;DR

This paper reviews various visualization techniques for analyzing computer vision datasets, highlighting their role in understanding data properties, predicting model impacts, and guiding dataset improvements.

Contribution

It provides a comprehensive survey of dataset visualization methods for computer vision, emphasizing their applications in dataset analysis and model performance prediction.

Findings

Visualization techniques reveal dataset properties and latent patterns.

Analysis informs mitigation strategies for dataset shortcomings.

Future directions include multimodal and task-specific visualization methods.

Abstract

We survey a number of data visualization techniques for analyzing Computer Vision (CV) datasets. These techniques help us understand properties and latent patterns in such data, by applying dataset-level analysis. We present various examples of how such analysis helps predict the potential impact of the dataset properties on CV models and informs appropriate mitigation of their shortcomings. Finally, we explore avenues for further visualization techniques of different modalities of CV datasets as well as ones that are tailored to support specific CV tasks and analysis needs.