Explaining 3D Semantic Segmentation Through Generative AI ‐Based Counterfactuals
Open Access
- 11 November 2025
- journal article
- Published by Wiley in Expert Systems
- Vol. 42 (12)
Abstract
Interpreting the predictions of deep learning models on 3D point cloud data is an important challenge for safety‐critical domains such as autonomous driving, robotics and geospatial analysis. Existing counterfactual explainability methods often struggle with the sparsity and unordered nature of 3D point clouds. To address this, we introduce a generative framework for counterfactual explanations in 3D semantic segmentation models. Our approach leverages autoencoder‐based latent representations, combined with UMAP embeddings and Delaunay triangulation, to construct a graph that enables geodesic path search between semantic classes. Candidate counterfactuals are generated by interpolating latent vectors along these paths and decoding into plausible point clouds, while semantic plausibility is guided by the predictions of a 3D semantic segmentation model. We evaluate the framework on ShapeNet objects, demonstrating that semantically related classes yield realistic counterfactuals with minimal geometric change, whereas unrelated classes expose sharp decision boundaries and reduced plausibility. Quantitative results confirm that the method balances defined interpretability metrics, producing counterfactuals that are both interpretable and geometrically consistent. Overall, our work demonstrates that generative counterfactuals in latent space provide a promising alternative to input‐level perturbations.Keywords
Funding Information
- Austrian Science Fund
- Eusko Jaurlaritza (IT1866‐26, KK‐2024/00064)
This publication has 36 references indexed in Scilit:
- Enhancing trust in automated 3D point cloud data interpretation through explainable counterfactualsInformation Fusion, 2025
- From 3D point‐cloud data to explainable geometric deep learning: State‐of‐the‐art and future challengesWIREs Data Mining and Knowledge Discovery, 2024
- Advancements in point cloud-based 3D defect classification and segmentation for industrial systems: A comprehensive surveyInformation Fusion, 2024
- A Perspective on Explainable Artificial Intelligence Methods: SHAP and LIMEAdvanced Intelligent Systems, 2024
- On generating trustworthy counterfactual explanationsInformation Sciences, 2023
- Beyond explaining: Opportunities and challenges of XAI-based model improvementInformation Fusion, 2022
- Towards multi-modal causability with Graph Neural Networks enabling information fusion for explainable AIInformation Fusion, 2021
- Multiresolution Tree Networks for 3D Point Cloud Processing2018
- Explaining nonlinear classification decisions with deep Taylor decompositionPattern Recognition, 2017
- When Is “Nearest Neighbor” Meaningful?1999