Statements (53)
| Predicate | Object |
|---|---|
| gptkbp:instance_of |
gptkb:cosmic_ray_detector
|
| gptkbp:application |
gptkb:medical_imaging
gptkb:robotics augmented reality autonomous driving image editing |
| gptkbp:applies_to |
computer vision tasks
|
| gptkbp:architecture |
encoder-decoder
|
| gptkbp:based_on |
Fully Convolutional Networks
|
| gptkbp:developed_by |
gptkb:Google_Research
|
| gptkbp:evaluates |
mean Intersection over Union (m Io U)
|
| gptkbp:feature_extraction |
gptkb:Res_Net
gptkb:Xception |
| gptkbp:field_of_study |
gptkb:Artificial_Intelligence
computer vision deep learning |
| gptkbp:has_achieved |
state-of-the-art performance
|
| https://www.w3.org/2000/01/rdf-schema#label |
Deep Labv1
|
| gptkbp:improves |
semantic segmentation
|
| gptkbp:input_output |
RGB images
segmentation maps |
| gptkbp:key_feature |
gptkb:Performance_Monitoring
community support high accuracy industrial applications real-time processing modular architecture contextual information research impact flexibility in design open-source availability support for various backbones dense prediction multi-scale context robustness to occlusion academic usage adaptability to different domains extensibility for future models transfer learning capability |
| gptkbp:losses |
cross-entropy loss
|
| gptkbp:performance |
gptkb:ADE20_K
gptkb:COCO_dataset |
| gptkbp:provides_information_on |
gptkb:PASCAL_VOC
Cityscapes |
| gptkbp:release_year |
gptkb:2016
|
| gptkbp:successor |
gptkb:Deep_Labv2
gptkb:Deep_Labv3 gptkb:Deep_Labv3+ |
| gptkbp:training |
supervised learning
|
| gptkbp:uses |
convolutional neural networks
|
| gptkbp:utilizes |
atrous convolution
|
| gptkbp:bfsParent |
gptkb:Deep_Lab
|
| gptkbp:bfsLayer |
5
|