Deep Residual Networks

GPTKB entity

Properties (52)
Predicate Object
gptkbp:instanceOf television channel
gptkbp:appearsIn Object Detection
Semantic Segmentation
gptkbp:basedOn Residual Learning
gptkbp:competesWith DenseNet
Inception_Networks
gptkbp:completed State-of-the-art_performance_on_ImageNet
gptkbp:developedBy gptkb:Kaiming_He
gptkbp:hasCitations Computationally intensive
Requires large datasets
gptkbp:hasImpactOn Computer_Vision
Deep_Learning_Research
gptkbp:hasRelatedPatent gptkb:Autonomous_Vehicles
Medical Imaging
Video Analysis
Facial_Recognition
gptkbp:hasVariants gptkb:ResNet-101
gptkb:ResNet-152
ResNet-50
https://www.w3.org/2000/01/rdf-schema#label Deep Residual Networks
gptkbp:improves Training of deep networks
gptkbp:influenced Subsequent Architectures
gptkbp:introduced 2015
gptkbp:isCharacterizedBy Identity Mapping
Layer-wise Learning
Deeper_Networks
gptkbp:isConsidered Standard Benchmark
Breakthrough_in_Deep_Learning
gptkbp:isCounteredBy gptkb:Adam_Optimizer
Stochastic Gradient Descent
gptkbp:isEvaluatedBy gptkb:ImageNet
gptkb:CIFAR-10
Top-1 Accuracy
Top-5 Accuracy
CIFAR-100
gptkbp:isOptimizedFor High Accuracy
Fast Training
gptkbp:isPopularIn Industry Applications
Research_Community
gptkbp:isRelatedTo gptkb:Convolutional_Neural_Networks
Transfer Learning
Feature Extraction
gptkbp:isSupportedBy Frameworks like TensorFlow
Frameworks_like_PyTorch
gptkbp:isUsedIn gptkb:ImageNet_Challenge
Research Papers
Kaggle Competitions
gptkbp:publishedIn CVPR 2016
gptkbp:usedFor Image Classification
gptkbp:uses Batch Normalization
ReLU Activation Function
gptkbp:utilizes Skip Connections