Cross-domain and Cross-dimension Learning for Image-to-Graph Transformers
Abstract
Direct image-to-graph transformation is a challenging task that involves solving object detection and relationship prediction in a single model. Due to this task's complexity, large training datasets are rare in many domains, making the training of deep-learning methods challenging. This data sparsity necessitates transfer learning strategies akin to the state-of-the-art in general computer vision. In this work, we introduce a set of methods enabling cross-domain and cross-dimension learning for image-to-graph transformers. We propose (1) a regularized edge sampling loss to effectively learn object relations in multiple domains with different numbers of edges, (2) a domain adaptation framework for image-to-graph transformers aligning image- and graph-level features from different domains, and (3) a projection function that allows using 2D data for training 3D transformers. We demonstrate our method's utility in cross-domain and cross-dimension experiments, where we utilize labeled data from 2D road networks for simultaneous learning in vastly different target domains. Our method consistently outperforms standard transfer learning and self-supervised pretraining on challenging benchmarks, such as retinal or whole-brain vessel graph extraction.
Cite
@article{arxiv.2403.06601,
title = {Cross-domain and Cross-dimension Learning for Image-to-Graph Transformers},
author = {Alexander H. Berger and Laurin Lux and Suprosanna Shit and Ivan Ezhov and Georgios Kaissis and Martin J. Menten and Daniel Rueckert and Johannes C. Paetzold},
journal= {arXiv preprint arXiv:2403.06601},
year = {2024}
}