SM 4 Depth: Seamless Monocular Metric Depth Estimation across Multiple Cameras and Scenes by One Model DOI
Yihao Liu, Feng Xue, Anlong Ming

et al.

Published: Oct. 26, 2024

Language: Английский

Exploiting Low-Level Representations for Ultra-Fast Road Segmentation DOI
Huan Zhou, Feng Xue, Yucong Li

et al.

IEEE Transactions on Intelligent Transportation Systems, Journal Year: 2024, Volume and Issue: 25(8), P. 9909 - 9919

Published: Aug. 1, 2024

Achieving real-time and accuracy on embedded platforms has always been the pursuit of road segmentation methods. To this end, they have proposed many lightweight networks. However, ignore fact that roads are "stuff" (background or environmental elements) rather than "things" (specific identifiable objects), which inspires us to explore feasibility representing with low-level instead high-level features. Surprisingly, we find primary stage mainstream network models is sufficient represent most pixels for segmentation. Motivated by this, propose a Low-level Feature Dominated Road Segmentation (LFD-RoadSeg). Specifically, LFD-RoadSeg employs bilateral structure. The spatial detail branch firstly designed extract feature representation first ResNet-18. suppress texture-less regions mistaken as in feature, context semantic then fast manner. second branch, asymmetrically downsample input image design an aggregation module achieve comparable receptive fields third ResNet-18 but less time consumption. Finally, segment from selective fusion calculate pixel-wise attention between non-road response attention. On KITTI-Road, achieves maximum F1-measure (MaxF) 95.21% average precision 93.71%, while reaching 238 FPS single TITAN Xp 54 Jetson TX2, all compact model size just 936k parameters. source code available at https://github.com/zhouhuan-hust/LFD-RoadSeg.

Language: Английский

Citations

2
Multi-View Remote Sensing Image Segmentation with Sam Priors DOI
Zipeng Qi, Chenyang Liu, Zili Liu

et al.

IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium, Journal Year: 2024, Volume and Issue: unknown, P. 8446 - 8449

Published: July 7, 2024

Language: Английский

Citations

1
Prompting Segment Anything Model with Domain-Adaptive Prototype for Generalizable Medical Image Segmentation DOI
Zhikai Wei, Wenhui Dong, Peilin Zhou

et al.

Lecture notes in computer science, Journal Year: 2024, Volume and Issue: unknown, P. 533 - 543

Published: Jan. 1, 2024

Language: Английский

Citations

0
EfficientMatting: Bilateral Matting Network for Real-Time Human Matting DOI

Rongsheng Luo,

Rukai Wei, Huaxin Zhang

et al.

Lecture notes in computer science, Journal Year: 2024, Volume and Issue: unknown, P. 128 - 142

Published: Nov. 2, 2024

Language: Английский

Citations

0
SM 4 Depth: Seamless Monocular Metric Depth Estimation across Multiple Cameras and Scenes by One Model DOI
Yihao Liu, Feng Xue, Anlong Ming

et al.

Published: Oct. 26, 2024

Language: Английский

Citations

0