Plant Phenomics / 2022 / Article / Tab 1

Research Article

PSegNet: Simultaneous Semantic and Instance Segmentation for Point Clouds of Plants

Table 1

Notations and nomenclatures.

FPSFarthest Point Sampling
VBSVoxelization-based Sampling
VFPSVoxelized Farthest Point Sampling
DNFEBDouble-Neighborhood Feature Extraction Block
DGFFMDouble-Granularity Feature Fusion Module
AMAttention Module
CAChannel attention
SASpatial attention
DHLDouble-hinge Loss
GTGround truth
MLPMultilayer perceptron
ReLURectified linear unit activation
PEPosition encoding
ECEdgeConv operation
APAttentive pooling

, Feature maps after decoding
Aggregated feature map after DGFFM
The loss functions
The number of semantic classes
The number of points in a point cloud
A point in space
A point vector in feature space
The parameter of KNN
Parameters for loss functions
Feature concatenation
The maximum value across the inputs
IoU of the two entities
MLP operation with shared parameters