TY - GEN
T1 - Anisotropic Multi-Scale Graph Convolutional Network for Dense Shape Correspondence
AU - Farazi, Mohammad
AU - Zhu, Wenhui
AU - Yang, Zhangsihao
AU - Wang, Yalin
N1 - Funding Information:
In the future, we will focus on the theoretical foundation of feature perturbation and explore how it can be generalized to, possibly, other existing models. Acknowledgment: The research is partly supported by NIH (R21AG065942, R01EY032125, R01EB025032, and R01DE030286).
Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - This paper studies 3D dense shape correspondence, a key shape analysis application in computer vision and graphics. We introduce a novel hybrid geometric deep learning-based model that learns geometrically meaningful and discretization-independent features. The proposed framework has a U-Net model as the primary node feature extractor, followed by a successive spectral-based graph convolutional network. To create a diverse set of filters, we use anisotropic wavelet basis filters, being sensitive to both different directions and band-passes. This filter set overcomes the common over-smoothing behavior of conventional graph neural networks. To further improve the model's performance, we add a function that perturbs the feature maps in the last layer ahead of fully connected layers, forcing the network to learn more discriminative features overall. The resulting correspondence maps show state-of-the-art performance on the benchmark datasets based on average geodesic errors and superior robustness to discretization in 3D meshes. Our approach provides new insights and practical solutions to the dense shape correspondence research.
AB - This paper studies 3D dense shape correspondence, a key shape analysis application in computer vision and graphics. We introduce a novel hybrid geometric deep learning-based model that learns geometrically meaningful and discretization-independent features. The proposed framework has a U-Net model as the primary node feature extractor, followed by a successive spectral-based graph convolutional network. To create a diverse set of filters, we use anisotropic wavelet basis filters, being sensitive to both different directions and band-passes. This filter set overcomes the common over-smoothing behavior of conventional graph neural networks. To further improve the model's performance, we add a function that perturbs the feature maps in the last layer ahead of fully connected layers, forcing the network to learn more discriminative features overall. The resulting correspondence maps show state-of-the-art performance on the benchmark datasets based on average geodesic errors and superior robustness to discretization in 3D meshes. Our approach provides new insights and practical solutions to the dense shape correspondence research.
KW - Algorithms: 3D computer vision
KW - Image recognition and understanding (object detection, categorization, segmentation, scene modeling, visual reasoning)
KW - Visualization
UR - http://www.scopus.com/inward/record.url?scp=85149006210&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85149006210&partnerID=8YFLogxK
U2 - 10.1109/WACV56688.2023.00316
DO - 10.1109/WACV56688.2023.00316
M3 - Conference contribution
AN - SCOPUS:85149006210
T3 - Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023
SP - 3145
EP - 3154
BT - Proceedings - 2023 IEEE Winter Conference on Applications of Computer Vision, WACV 2023
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 23rd IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2023
Y2 - 3 January 2023 through 7 January 2023
ER -