TY - GEN
T1 - VIASEG
T2 - 26th IEEE International Conference on Image Processing, ICIP 2019
AU - Zhong, Zhibin
AU - Zhang, Chi
AU - Liu, Yuehu
AU - Wu, Ying
N1 - Funding Information:
★This work is supported by National Nature Science Foundation of China (Grant No: 91520301).
Funding Information:
This work is supported by National Nature Science Foundation of China (Grant No: 91520301).
Publisher Copyright:
© 2019 IEEE.
PY - 2019/9
Y1 - 2019/9
N2 - Rapid and precise point cloud segmentation is one of the prerequisites for real-time and robust autonomous perception and environmental understanding, which requires a balance between speed and accuracy in architecture design. However, recent lightweight architectures, though fast enough, rely on domain adaptation from time-consuming-constructed synthetic dataset and sophisticated post-processing procedure to improve their performance, neglecting the rich visual information acquired by cameras aside from LiDAR sensors. In this paper, such color information is embedded at data-level to boost the performance of real-time point cloud segmentation. Furthermore, a multiscale lightweight fully convolutional network, VIASeg, is proposed based on the newly designed Super Squeeze Residual module and Semantic Connection from higher convolutional layers to lower layers, which improves the performance by feature denoising with high level semantic information. The superiority of the proposed method is validated and demonstrated in the comparative and ablative experimental analysis, while maintaining the real-time characteristic.
AB - Rapid and precise point cloud segmentation is one of the prerequisites for real-time and robust autonomous perception and environmental understanding, which requires a balance between speed and accuracy in architecture design. However, recent lightweight architectures, though fast enough, rely on domain adaptation from time-consuming-constructed synthetic dataset and sophisticated post-processing procedure to improve their performance, neglecting the rich visual information acquired by cameras aside from LiDAR sensors. In this paper, such color information is embedded at data-level to boost the performance of real-time point cloud segmentation. Furthermore, a multiscale lightweight fully convolutional network, VIASeg, is proposed based on the newly designed Super Squeeze Residual module and Semantic Connection from higher convolutional layers to lower layers, which improves the performance by feature denoising with high level semantic information. The superiority of the proposed method is validated and demonstrated in the comparative and ablative experimental analysis, while maintaining the real-time characteristic.
KW - Cross-modality Fusion
KW - Fully Convolutional Residual Network
KW - Point Cloud Segmentation
UR - http://www.scopus.com/inward/record.url?scp=85076823235&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85076823235&partnerID=8YFLogxK
U2 - 10.1109/ICIP.2019.8803061
DO - 10.1109/ICIP.2019.8803061
M3 - Conference contribution
AN - SCOPUS:85076823235
T3 - Proceedings - International Conference on Image Processing, ICIP
SP - 1500
EP - 1504
BT - 2019 IEEE International Conference on Image Processing, ICIP 2019 - Proceedings
PB - IEEE Computer Society
Y2 - 22 September 2019 through 25 September 2019
ER -