Statements (62)
| Predicate | Object |
|---|---|
| gptkbp:instance_of |
gptkb:neural_networks
|
| gptkbp:achieves_top1_accuracy |
74.9% on Image Net
|
| gptkbp:achieves_top5_accuracy |
91.2% on Image Net
|
| gptkbp:based_on |
neural architecture search
|
| gptkbp:can_be_used_for |
object detection
image generation semantic segmentation |
| gptkbp:competes_with |
gptkb:Dense_Net
gptkb:Res_Net |
| gptkbp:designed_for |
image classification
|
| gptkbp:developed_by |
gptkb:Google
|
| gptkbp:has_achieved |
state-of-the-art performance
|
| gptkbp:has_function |
88 million
|
| gptkbp:has_variants |
gptkb:NASNet-A
gptkb:NASNet-C NASNet-D |
| https://www.w3.org/2000/01/rdf-schema#label |
NASNet-B
|
| gptkbp:introduced_in |
gptkb:2018
|
| gptkbp:is_available_in |
gptkb:Tensor_Flow
gptkb:Keras gptkb:Py_Torch |
| gptkbp:is_documented_in |
research articles
technical reports theses conference papers Git Hub repositories |
| gptkbp:is_evaluated_by |
gptkb:CIFAR-10
gptkb:Stanford_Dogs gptkb:CIFAR-100 F1 score mean average precision Top-1 accuracy Top-5 accuracy Oxford Pets |
| gptkbp:is_influenced_by |
human-designed architectures
|
| gptkbp:is_optimized_for |
GPU acceleration
|
| gptkbp:is_part_of |
gptkb:machine_learning
|
| gptkbp:is_related_to |
hyperparameter tuning
ensemble methods adversarial training model compression techniques transfer learning techniques data augmentation methods |
| gptkbp:is_supported_by |
gptkb:academic_research
community contributions open-source projects |
| gptkbp:is_trained_in |
gptkb:Image_Net
large datasets |
| gptkbp:is_used_in |
research papers
industry applications |
| gptkbp:part_of |
NASNet family
|
| gptkbp:performance |
neural architecture search
deep learning models image recognition tasks computer vision tasks |
| gptkbp:supports |
transfer learning
|
| gptkbp:uses |
batch normalization
convolutional layers dropout layers recurrent layers |
| gptkbp:bfsParent |
gptkb:NASNet
|
| gptkbp:bfsLayer |
6
|