A Framework for Evaluating Zero-Shot Image Generation in Concept-based Explainability

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Astolfi, Giacomo, Bianchi, Matteo, Campi, Riccardo, De Santis, Antonio, Brambilla, Marco
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916028002336768
author Astolfi, Giacomo
Bianchi, Matteo
Campi, Riccardo
De Santis, Antonio
Brambilla, Marco
author_facet Astolfi, Giacomo
Bianchi, Matteo
Campi, Riccardo
De Santis, Antonio
Brambilla, Marco
contents Concept-based Explainable Artificial Intelligence (XAI) interprets deep learning models using human-understandable visual features (e.g., textures or object parts) by linking internal representations to class predictions, thereby bridging the gap between low-level image data and high-level semantics. A major challenge, however, is the reliance on large sets of labeled images to represent each concept, which limits scalability. In this work, we investigate the use of zero-shot Text-to-Image (T2I) generative models as a source of synthetic concept datasets for concept-based XAI methods. Specifically, we generate concepts using predefined prompts and evaluate their faithfulness to real ones through four complementary analyses: (1) comparing synthetic vs. real concept images via concept representation similarity; (2) evaluating their intra-similarity by comparing pairs of subsets of the same concept with progressively increasing size; (3) evaluating their performance for downstream explanation tasks using relevant class images; (4) evaluating how removing a concept from tested class images affects explanations of generated concepts. While current T2I generative models promise a shortcut to concept-based XAI, our study highlights challenges and raises open questions about the use of synthetic data generated by zero-shot pipelines in model analyses. The resulting dataset is available at https://github.com/DataSciencePolimi/ZeroShot-T2I-Concepts.
format Preprint
id arxiv_https___arxiv_org_abs_2605_19855
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle A Framework for Evaluating Zero-Shot Image Generation in Concept-based Explainability
Astolfi, Giacomo
Bianchi, Matteo
Campi, Riccardo
De Santis, Antonio
Brambilla, Marco
Computer Vision and Pattern Recognition
Artificial Intelligence
Concept-based Explainable Artificial Intelligence (XAI) interprets deep learning models using human-understandable visual features (e.g., textures or object parts) by linking internal representations to class predictions, thereby bridging the gap between low-level image data and high-level semantics. A major challenge, however, is the reliance on large sets of labeled images to represent each concept, which limits scalability. In this work, we investigate the use of zero-shot Text-to-Image (T2I) generative models as a source of synthetic concept datasets for concept-based XAI methods. Specifically, we generate concepts using predefined prompts and evaluate their faithfulness to real ones through four complementary analyses: (1) comparing synthetic vs. real concept images via concept representation similarity; (2) evaluating their intra-similarity by comparing pairs of subsets of the same concept with progressively increasing size; (3) evaluating their performance for downstream explanation tasks using relevant class images; (4) evaluating how removing a concept from tested class images affects explanations of generated concepts. While current T2I generative models promise a shortcut to concept-based XAI, our study highlights challenges and raises open questions about the use of synthetic data generated by zero-shot pipelines in model analyses. The resulting dataset is available at https://github.com/DataSciencePolimi/ZeroShot-T2I-Concepts.
title A Framework for Evaluating Zero-Shot Image Generation in Concept-based Explainability
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2605.19855