AEye: A Visualization Tool for Image Datasets
This tool addresses the need for intuitive exploration of image datasets for researchers and practitioners in computer vision, but it is incremental as it builds on existing visualization and embedding techniques.
The authors tackled the problem of understanding image dataset composition and distribution by developing AEye, an extensible visualization tool that embeds images into semantic representations for clustering and interactive exploration, resulting in a tool that supports semantic search for text and image queries and is open-sourced with simple dataset configuration.
Image datasets serve as the foundation for machine learning models in computer vision, significantly influencing model capabilities, performance, and biases alongside architectural considerations. Therefore, understanding the composition and distribution of these datasets has become increasingly crucial. To address the need for intuitive exploration of these datasets, we propose AEye, an extensible and scalable visualization tool tailored to image datasets. AEye utilizes a contrastively trained model to embed images into semantically meaningful high-dimensional representations, facilitating data clustering and organization. To visualize the high-dimensional representations, we project them onto a two-dimensional plane and arrange images in layers so users can seamlessly navigate and explore them interactively. AEye facilitates semantic search functionalities for both text and image queries, enabling users to search for content. We open-source the codebase for AEye, and provide a simple configuration to add datasets.