Evaluating the Robustness and Generalizability of D-LeDe Using Multiple Automotive Datasets
SN Computer Science · 20.08.2026 · 7 (7)
DOI: 10.1007/s42979-026-05289-7 ↗Kurz gefasst
Fachartikel in der Zeitschrift SN Computer Science, online erschienen am 20.08.2026. Verfasst von Md Abu Ahammed Babu und 4 weiteren Autor:innen. Die Arbeit nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Chalmers University of Technology.
Abstract (Originalsprache)
Data leakage, where semantically or visually similar samples exist across training and test splits, continues to threaten the reliability of object detection benchmarks. The D-LeDe method was recently proposed as a statistical technique for detecting data leakage, showing promise in initial applications. This paper extends the D-LeDe method to evaluate its robustness and generalizability through a focused experimental study. We revisit the KITTI dataset which was previously identified as leakage-prone and introduce a new application of D-LeDe on SODA10M. The core investigation centers on whether visually similar images, measured using perceptual hashing, are the primary cause of data leakage indications captured by D-LeDe. To this end, we progressively remove visually similar image pairs from the test sets of both datasets and observe changes in the Relative Increase Rate, the key decision metric of D-LeDe. Results show that even after removing about 50% of similar images from the KITTI test set, D-LeDe continues to detect data leakage, suggesting other forms of redundancy or latent factors causing the leakage. In contrast, SODA10M consistently remains leakage-free across all levels of image pair removal. These findings reinforce the reliability of D-LeDe in detecting non-obvious forms of data leakage, underscoring its applicability as a diagnostic tool for dataset integrity in object detection workflows.
Förderung
- Chalmers University of Technology
Zitieren
APA
Babu, M. A. A., Staron, M., Durisic, D., Bálint, A., & Pandey, S. K. (2026). Evaluating the Robustness and Generalizability of D-LeDe Using Multiple Automotive Datasets. SN Computer Science, 7(7). https://doi.org/10.1007/s42979-026-05289-7
Vancouver
Babu MAA, Staron M, Durisic D, Bálint A, Pandey SK. Evaluating the Robustness and Generalizability of D-LeDe Using Multiple Automotive Datasets. SN Computer Science. 2026;7(7). doi:10.1007/s42979-026-05289-7
BibTeX
@article{Babu2026,
author = {Babu, Md Abu Ahammed and Staron, Miroslaw and Durisic, Darko and Bálint, András and Pandey, Sushant Kumar},
title = {Evaluating the Robustness and Generalizability of D-LeDe Using Multiple Automotive Datasets},
journal = {SN Computer Science},
year = {2026},
volume = {7},
number = {7},
doi = {10.1007/s42979-026-05289-7}
}
EUVIMED