Statements (53)
| Predicate | Object |
|---|---|
| gptkbp:instance_of |
gptkb:television_channel
|
| gptkbp:bfsLayer |
4
|
| gptkbp:bfsParent |
gptkb:Dense_Net
|
| gptkbp:applies_to |
Image Classification
|
| gptkbp:architectural_style |
gptkb:Deep_Learning
|
| gptkbp:characteristics |
gptkb:Community_Center
Scalability Global Average Pooling Model Interpretability Robustness to Noise Growth Rate State-of-the-Art Performance Reduced Overfitting Compression Factor Research Interest No Fully Connected Layers Gradient Flow Bottleneck Layers Compact Model Size Efficient Parameter Usage Feature Reuse High Feature Extraction Capability High Performance on Small Datasets Hyperparameter Tuning Required Improved Training Speed Layer Connectivity Multi-Scale Feature Learning Open Source Implementation Transfer Learning Capability Transition Layers Visualization Techniques Available Widely Used in Competitions |
| gptkbp:coat_of_arms |
169
|
| gptkbp:developed_by |
Gao Huang
|
| gptkbp:has_achievements |
High Accuracy
|
| https://www.w3.org/2000/01/rdf-schema#label |
Dense Net-169
|
| gptkbp:inspired_by |
gptkb:Res_Net
|
| gptkbp:is_a_framework_for |
gptkb:Graphics_Processing_Unit
gptkb:Keras gptkb:Py_Torch |
| gptkbp:is_compared_to |
gptkb:Res_Net
|
| gptkbp:performance |
Top-1 Accuracy
Top-5 Accuracy |
| gptkbp:predecessor |
gptkb:Dense_Net-121
|
| gptkbp:provides_information_on |
gptkb:Image_Net
|
| gptkbp:requires |
GPU for Training
|
| gptkbp:successor |
gptkb:Dense_Net-201
|
| gptkbp:uses |
gptkb:Batch_Normalization
gptkb:Dropout Dense Connectivity Re LU Activation Function |
| gptkbp:written_in |
gptkb:Library
|
| gptkbp:year_created |
gptkb:2017
|