CAD Models to Real-World Images: A Practical Approach to Unsupervised Domain Adaptation in Industrial Object Classification

Dennis Ritter1, Mike Hemberger2, Marc Hönig3, Volker Stopp3, Erik Rodner4, Kristian Hildebrand1,
1Berliner Hochschule für Technik, 2nyris GmbH, 3topex GmbH, 4KI-Werkstatt/FB2 University of Applied Sciences Berlin
ECML-PKDD 2023 - Adapting to Change: Reliable Multimodal Learning Across Domains
Paper Transfer Learning Framework Render Pipeline Dataset
Gefördert vom Bundesministerium für Bildung und Forschung mit Logo

Förderkennzeichen: 01IS21002C

Abstract

In this paper, we systematically analyze unsupervised domain adaptation pipelines for object classification in a challenging industrial setting. In contrast to standard natural object benchmarks existing in the field, our results highlight the most important design choices when only category-labeled CAD models are available but classification needs to be done with real-world images. Our domain adaptation pipeline achieves SoTA performance on the VisDA benchmark, but more importantly, drastically improves recognition performance on our new open industrial dataset comprised of 102 mechanical parts. We conclude with a set of guidelines that are relevant for practitioners needing to apply state-of-the-art unsupervised domain adaptation in practice.

Random examples of our rendered dataset

Render Results

Random examples of our real photo dataset

Test Dataset containing only real images

Overview of our render pipeline

Render pipeline diagram

Overview of our training method

Training method diagram

Poster

BibTeX

@InProceedings{10.1007/978-3-031-74640-6_33,
                        author="Ritter, Dennis
                        and Hemberger, Mike
                        and H{\"o}nig, Marc
                        and Stopp, Volker
                        and Rodner, Erik
                        and Hildebrand, Kristian",
                        editor="Meo, Rosa
                        and Silvestri, Fabrizio",
                        title="CAD Models to Real-World Images: A Practical Approach to Unsupervised Domain Adaptation in Industrial Object Classification",
                        booktitle="Machine Learning and Principles and Practice of Knowledge Discovery in Databases",
                        year="2025",
                        publisher="Springer Nature Switzerland",
                        address="Cham",
                        pages="399--415",
                        abstract="In this paper, we systematically analyze unsupervised domain adaptation pipelines for object classification in a challenging industrial setting. In contrast to standard natural object benchmarks existing in the field, our results highlight the most important design choices when only category-labeled CAD models are available but classification needs to be done with real-world images. Our domain adaptation pipeline achieves SoTA performance on the VisDA benchmark, but more importantly, drastically improves recognition performance on our new open industrial dataset comprised of 102 mechanical parts. We conclude with a set of guidelines that are relevant for practitioners needing to apply state-of-the-art unsupervised domain adaptation in practice. Our code is available at https://github.com/dritter-bht/synthnet-transfer-learning.",
                        isbn="978-3-031-74640-6"
                        }