RetinaNet

E431010

RetinaNet is a deep learning–based one-stage object detection model known for its focal loss function, which effectively addresses class imbalance to achieve high accuracy and speed.

All labels observed (2)

Label Occurrences
RetinaNet canonical 2
Focal Loss for Dense Object Detection 1

How this entity was disambiguated

Statements (49)

Predicate Object
instanceOf convolutional neural network
deep learning model
object detection model
one-stage detector
accuracyCharacteristic high detection accuracy
addressesProblem class imbalance in object detection
foreground-background class imbalance
basedOn cross-entropy loss
category computer vision model
codeAvailability open source implementations exist
comparedTo two-stage detectors
designedFor dense object detection
detectionStageType single-stage
evaluatedOn COCO dataset
field computer vision
deep learning
hasArchitectureComponent box regression subnet
classification subnet
hasAuthor Kaiming He
Piotr Dollár
Priya Goyal
Ross Girshick
Tsung-Yi Lin
hasBackbone ResNet
ResNet-101
linked to: ResNet

ResNet-50
linked to: ResNet
hasHyperparameter alpha in focal loss
gamma in focal loss
hasKeyConcept focal loss
implementationFramework Caffe2
PyTorch
TensorFlow
inputType natural images
introducedInPaper Focal Loss for Dense Object Detection
linked to: RetinaNet
lossFunctionProperty down-weights easy examples
focuses on hard, misclassified examples
optimizationObjective focal loss
outperforms many two-stage detectors on COCO
outputType bounding boxes
class probabilities
publicationYear 2017
speedCharacteristic high inference speed
task object detection in images
trainingDataset MS COCO
linked to: COCO
trainingType supervised learning
usesAnchorBoxes true
usesComponent FPN
Feature Pyramid Network
linked to: FPN
usesMultiScaleFeatures true

How these facts were elicited

Referenced by (3)

Full triples — surface form annotated when it differs from this entity's canonical label.

torchvision modelFamily RetinaNet
subject linked to: torchvision (ecosystem)
RetinaNet introducedInPaper Focal Loss for Dense Object Detection
linked to: RetinaNet
DETR comparedTo RetinaNet