Please enable JavaScript.
Coggle requires JavaScript to display documents.
YOLOv4 - Coggle Diagram
YOLOv4
Input : Data Augumentation
Mosaic data augmentation
CutMix
Class label smoothing
Self-Adversarial Training
Loss
Focal loss
OHEM
IOU
GIOU
DIOU
CIOU
Regularization
DropOut
DropConnect
DropBlock
Neck(feature enhancement module)
Filter
PANet
ASPP
RFB
SPP
Feature Fusion
FPN
SFAM
ASFF
BiFPN
Backbones
GPU
CSPResNeXt50 / CSPDarknet53
VPU
EfficientNet-lite
MixNet
GhostNet
MobileNetV3
Attention
Squeeze-and-Excitation(SE)
Spatial Attention Module(SAM)
NMS
Soft-NMS
DIoU NMS
Activation
Swish
hard-Swish
Head
YoloV3