|
gptkbp:instance_of
|
gptkb:television_channel
|
|
gptkbp:applies_to
|
autonomous driving
medical image analysis
satellite image processing
|
|
gptkbp:architectural_style
|
gptkb:television_channel
|
|
gptkbp:based_on
|
gptkb:Alex_Net
|
|
gptkbp:competes_with
|
gptkb:Alumni_Association
|
|
gptkbp:developed_by
|
gptkb:Jonathan_Long
|
|
gptkbp:enhances
|
image understanding
|
|
gptkbp:has
|
encoder-decoder structure
|
|
gptkbp:has_achievements
|
end-to-end training
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
FCN-16s
|
|
gptkbp:improves
|
object localization
pixel-wise classification
|
|
gptkbp:input_output
|
gptkb:Photographer
segmentation map
class probabilities
|
|
gptkbp:is_compared_to
|
gptkb:Deep_Lab
gptkb:Mask_R-CNN
gptkb:Seg_Net
|
|
gptkbp:is_designed_for
|
semantic segmentation
|
|
gptkbp:is_evaluated_by
|
benchmark datasets
mean Intersection over Union (m Io U)
|
|
gptkbp:is_implemented_in
|
gptkb:Graphics_Processing_Unit
gptkb:Py_Torch
|
|
gptkbp:is_influenced_by
|
CNN architectures
|
|
gptkbp:is_known_for
|
pixel-level predictions
|
|
gptkbp:is_optimized_for
|
speed and accuracy
|
|
gptkbp:is_related_to
|
image processing
computer vision
deep learning
|
|
gptkbp:is_used_in
|
real-time applications
|
|
gptkbp:requires
|
large datasets
|
|
gptkbp:training
|
gptkb:Pascal_VOC
gptkb:COCO
Cityscapes
|
|
gptkbp:uses
|
feature maps
skip connections
deconvolution layers
|
|
gptkbp:utilizes
|
backpropagation
|
|
gptkbp:year_created
|
gptkb:2015
|
|
gptkbp:bfsParent
|
gptkb:FCN-8
|
|
gptkbp:bfsLayer
|
5
|