A Cascade Dense Connection Fusion Network for Depth Completion
Rizhao Fan (University of Bologna),* Zhigen Li (Ping An Technology), Matteo Poggi (University of Bologna), Stefano Mattoccia (University of Bologna)
The 33rd British Machine Vision Conference
Abstract
This paper proposes a lightweight yet effective cascade-guided network architecture for depth completion. It enables to fuse multi-modal and multi-level features through a lightweight Cascade Dense Connection Fusion Network. Moreover, we design a dense connection fusion block that adopts dense connections, multi-scale learning and a modality-aware mechanism. Our model is evaluated on the established KITTI benchmark and achieves competitive results compared with state-of-the-art while counting fewer parameters.Video
Citation
@inproceedings{Fan_2022_BMVC,
author = {Rizhao Fan and Zhigen Li and Matteo Poggi and Stefano Mattoccia},
title = {A Cascade Dense Connection Fusion Network for Depth Completion},
booktitle = {33rd British Machine Vision Conference 2022, {BMVC} 2022, London, UK, November 21-24, 2022},
publisher = {{BMVA} Press},
year = {2022},
url = {https://bmvc2022.mpi-inf.mpg.de/0843.pdf}
}
Copyright © 2022
The British Machine Vision Conference is organised by