Backbone deep learning là gì
According to my understanding, the "backbone" refers to the DeepLab architecture's feature extracting network. The network's input is encoded into a feature representation using this feature extractor. Around this feature extractor, the DeepLab framework "wraps" functionalities. The feature extractor can then be swapped out, and a model can be picked that best suits the task at hand in terms of accuracy, efficiency, and so on. Show
Bag of Freebies vs Bag of Specials
BackboneDense Block & DenseNet
Cross-stage-partial-connection (CSP)
CSPDarknet53
Neck
Feature Pyramid Networks (FPN)SPP (spatial pyramid pooling layer)YOLO with SPP
Path Aggregation Network (PAN) – ref
Spatial Attention Module (SAM)
Yolov4 sử dụng gì?Bag of Freebies cho backbone
1. CutMix data augmentation
2. Mosaic data augmentation
3. DropBlock regularization
4. Class label smoothing
Bag of Specials (BoS) cho backbone
1. Mish activation
From the paper- a comparison of the output landscape from ReLU, Swish and Mish. The smooth gradients from Mish is a likely driver of it’s outperformance.
2. Cross-stage partial connections (CSP) – [Xem bên trên]3. Multi-input weighted residual connections (MiWRC)
Bag of Freebies (BoF) for detector
1. CIoU-loss
2. CmBN
3. DropBlock regularization 4. Mosaic data augmentation 5. Self-Adversarial Training
6. Eliminate grid sensitivity
7. Using multiple anchors for a single ground truth
8. Cosine annealing scheduler
9. Optimal hyperparameters
10. Random training shapes
Bag of Specials (BoS) for detector
Reference
|