|
gptkbp:instance_of
|
gptkb:neural_networks
|
|
gptkbp:application
|
Image classification
|
|
gptkbp:architecture
|
gptkb:Efficient_Net
|
|
gptkbp:based_on
|
Compound scaling
|
|
gptkbp:dataset_trained_on
|
gptkb:Image_Net
|
|
gptkbp:designed_for
|
Mobile and edge devices
|
|
gptkbp:developed_by
|
gptkb:Google_AI
|
|
gptkbp:focus_area
|
gptkb:True
|
|
gptkbp:has_achieved
|
State-of-the-art results
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
Efficient Net B6
|
|
gptkbp:influenced_by
|
gptkb:Dense_Net
gptkb:Res_Net
|
|
gptkbp:input_output
|
528x528
|
|
gptkbp:is_a_framework_for
|
gptkb:Tensor_Flow
|
|
gptkbp:is_compared_to
|
gptkb:Efficient_Net_B0
gptkb:Efficient_Net_B1
gptkb:Efficient_Net_B2
gptkb:Efficient_Net_B3
gptkb:Efficient_Net_B4
gptkb:Efficient_Net_B5
|
|
gptkbp:is_optimized_for
|
Accuracy and efficiency
|
|
gptkbp:license
|
Apache License 2.0
|
|
gptkbp:notable_feature
|
Efficient scaling of depth, width, and resolution
|
|
gptkbp:orbital_period
|
Billion
|
|
gptkbp:performance
|
Top-1 accuracy
|
|
gptkbp:pretrained_models_available
|
gptkb:Yes
|
|
gptkbp:release_year
|
gptkb:2020
|
|
gptkbp:repository
|
gptkb:Git_Hub
|
|
gptkbp:successor
|
gptkb:Efficient_Net_B7
|
|
gptkbp:supports_multi_gputraining
|
gptkb:Yes
|
|
gptkbp:supports_transfer_learning
|
gptkb:Yes
|
|
gptkbp:top1_accuracy
|
84.4%
|
|
gptkbp:top5_accuracy
|
97.1%
|
|
gptkbp:tuning
|
gptkb:True
|
|
gptkbp:uses
|
Swish activation function
|
|
gptkbp:uses_adam_optimizer
|
gptkb:True
|
|
gptkbp:uses_aug_mix
|
gptkb:True
|
|
gptkbp:uses_auto_augment
|
gptkb:True
|
|
gptkbp:uses_batch_normalization
|
gptkb:True
|
|
gptkbp:uses_cross_validation
|
gptkb:True
|
|
gptkbp:uses_cut_mix_training
|
gptkb:True
|
|
gptkbp:uses_cutout
|
gptkb:True
|
|
gptkbp:uses_data_augmentation
|
gptkb:True
|
|
gptkbp:uses_drop_connect
|
gptkb:True
|
|
gptkbp:uses_dropout
|
gptkb:True
|
|
gptkbp:uses_early_stopping
|
gptkb:True
|
|
gptkbp:uses_ensemble_methods
|
gptkb:True
|
|
gptkbp:uses_generative_adversarial_networks
|
gptkb:True
|
|
gptkbp:uses_global_average_pooling
|
gptkb:True
|
|
gptkbp:uses_gradient_clipping
|
gptkb:True
|
|
gptkbp:uses_hyperparameter_tuning
|
gptkb:True
|
|
gptkbp:uses_image_net_preprocessing
|
gptkb:True
|
|
gptkbp:uses_keras
|
gptkb:True
|
|
gptkbp:uses_l2_regularization
|
gptkb:True
|
|
gptkbp:uses_label_smoothing
|
gptkb:True
|
|
gptkbp:uses_learning_rate_scheduler
|
gptkb:True
|
|
gptkbp:uses_mixed_precision_training
|
gptkb:True
|
|
gptkbp:uses_mixup_training
|
gptkb:True
|
|
gptkbp:uses_model_averaging
|
gptkb:True
|
|
gptkbp:uses_model_checkpointing
|
gptkb:True
|
|
gptkbp:uses_neural_architecture_search
|
gptkb:True
|
|
gptkbp:uses_rand_augment
|
gptkb:True
|
|
gptkbp:uses_random_erasing
|
gptkb:True
|
|
gptkbp:uses_reinforcement_learning
|
gptkb:True
|
|
gptkbp:uses_self_supervised_learning
|
gptkb:True
|
|
gptkbp:uses_semi_supervised_learning
|
gptkb:True
|
|
gptkbp:uses_squeeze_excitation
|
gptkb:True
|
|
gptkbp:uses_tensor_board
|
gptkb:True
|
|
gptkbp:uses_tfrecord_format
|
gptkb:True
|
|
gptkbp:uses_transfer_learning
|
gptkb:True
|
|
gptkbp:uses_unsupervised_learning
|
gptkb:True
|
|
gptkbp:uses_variational_autoencoders
|
gptkb:True
|
|
gptkbp:bfsParent
|
gptkb:Efficient_Net
|
|
gptkbp:bfsLayer
|
5
|