TY - GEN
T1 - VMNet
T2 - 18th IEEE/CVF International Conference on Computer Vision, ICCV 2021
AU - Hu, Zeyu
AU - Bai, Xuyang
AU - Shang, Jiaxiang
AU - Zhang, Runze
AU - Dong, Jiayu
AU - Wang, Xin
AU - Sun, Guangyuan
AU - Fu, Hongbo
AU - Tai, Chiew-Lan
N1 - Full text of this publication does not contain sufficient affiliation information. With consent from the author(s) concerned, the Research Unit(s) information for this record is based on the existing academic department affiliation of the author(s).
PY - 2021/10
Y1 - 2021/10
N2 - In recent years, sparse voxel-based methods have become the state-of-the-arts for 3D semantic segmentation of indoor scenes, thanks to the powerful 3D CNNs. Nevertheless, being oblivious to the underlying geometry, voxel-based methods suffer from ambiguous features on spatially close objects and struggle with handling complex and irregular geometries due to the lack of geodesic information. In view of this, we present Voxel-Mesh Network (VMNet), a novel 3D deep architecture that operates on the voxel and mesh representations leveraging both the Euclidean and geodesic information. Intuitively, the Euclidean information extracted from voxels can offer contextual cues representing interactions between nearby objects, while the geodesic information extracted from meshes can help separate objects that are spatially close but have disconnected surfaces. To incorporate such information from the two domains, we design an intra-domain attentive module for effective feature aggregation and an inter-domain attentive module for adaptive feature fusion. Experimental results validate the effectiveness of VMNet: specifically, on the challenging ScanNet dataset for large-scale segmentation of indoor scenes, it outperforms the state-of-the-art SparseConvNet and MinkowskiNet (74.6% vs 72.5% and 73.6% in mIoU) with a simpler network structure (17M vs 30M and 38M parameters). Code release: https://github.com/hzykent/VMNet.
AB - In recent years, sparse voxel-based methods have become the state-of-the-arts for 3D semantic segmentation of indoor scenes, thanks to the powerful 3D CNNs. Nevertheless, being oblivious to the underlying geometry, voxel-based methods suffer from ambiguous features on spatially close objects and struggle with handling complex and irregular geometries due to the lack of geodesic information. In view of this, we present Voxel-Mesh Network (VMNet), a novel 3D deep architecture that operates on the voxel and mesh representations leveraging both the Euclidean and geodesic information. Intuitively, the Euclidean information extracted from voxels can offer contextual cues representing interactions between nearby objects, while the geodesic information extracted from meshes can help separate objects that are spatially close but have disconnected surfaces. To incorporate such information from the two domains, we design an intra-domain attentive module for effective feature aggregation and an inter-domain attentive module for adaptive feature fusion. Experimental results validate the effectiveness of VMNet: specifically, on the challenging ScanNet dataset for large-scale segmentation of indoor scenes, it outperforms the state-of-the-art SparseConvNet and MinkowskiNet (74.6% vs 72.5% and 73.6% in mIoU) with a simpler network structure (17M vs 30M and 38M parameters). Code release: https://github.com/hzykent/VMNet.
UR - https://www.scopus.com/pages/publications/85120762000
UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85120762000&origin=recordpage
U2 - 10.1109/ICCV48922.2021.01520
DO - 10.1109/ICCV48922.2021.01520
M3 - RGC 32 - Refereed conference paper (with host publication)
T3 - Proceedings of the IEEE International Conference on Computer Vision
SP - 15468
EP - 15474
BT - Proceedings - 2021 IEEE/CVF International Conference on Computer Vision (ICCV 2021)
PB - IEEE
Y2 - 11 October 2021 through 17 October 2021
ER -