Monocular depth estimation with spatially coherent sliced network
作者：
Highlights：
• Spatially coherent sliced network regresses sliced depth from feature fusions pyramidally.
• Our network introduces the scale and shift regression to maintain spatial coherence.
• Self-spatial-attention mechanism refine potentially compromised semantic coherence.
• Experimental results show competitive results and promising performance in urban road scenes.
摘要
Highlights•Spatially coherent sliced network regresses sliced depth from feature fusions pyramidally.•Our network introduces the scale and shift regression to maintain spatial coherence.•Self-spatial-attention mechanism refine potentially compromised semantic coherence.•Experimental results show competitive results and promising performance in urban road scenes.
论文关键词：Depth estimation,Monocular images,Spatial coherence,Sliced depth
论文评审过程：Received 2 July 2021, Revised 14 May 2022, Accepted 16 May 2022, Available online 30 May 2022, Version of Record 17 June 2022.
论文官网地址：https://doi.org/10.1016/j.imavis.2022.104487
