Sparse Gaussian process for online seagrass semantic mapping
作者：
Highlights：
• Ease the generation of seagrass data on the spatial distribution.
• Use encoder–decoder CNN to segment benthic images online.
• Use sparse GP to map uncertainty.
• Complete assessment of different sparse GPs and CNNs configurations.
• Set of experiments conducted in marine scenarios colonised with Posidonia Oceanica.
摘要
•Ease the generation of seagrass data on the spatial distribution.•Use encoder–decoder CNN to segment benthic images online.•Use sparse GP to map uncertainty.•Complete assessment of different sparse GPs and CNNs configurations.•Set of experiments conducted in marine scenarios colonised with Posidonia Oceanica.
论文关键词：Robotics,AUV,Semantic mapping,Seagrass,Posidonia oceanica,Gaussian processes
论文评审过程：Received 16 January 2020, Revised 5 December 2020, Accepted 6 December 2020, Available online 24 December 2020, Version of Record 7 January 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2020.114478
