Training Techniques | Polynomial Learning Rate Decay, Self-Adversarial Training, Weight Decay, SGD with Momentum, Label Smoothing, Mosaic, CutMix |
---|---|
Architecture | 1x1 Convolution, Batch Normalization, Convolution, Global Average Pooling, Mish, Residual Connection, Softmax |
ID | cspdarknet53 |
SHOW MORE |
CSPDarknet53 is a convolutional neural network and backbone for object detection that uses DarkNet-53. It employs a CSPNet strategy to partition the feature map of the base layer into two parts and then merges them through a cross-stage hierarchy. The use of a split and merge strategy allows for more gradient flow through the network.
This CNN is used as the backbone for YOLOv4.
To load a pretrained model:
import timm
m = timm.create_model('cspdarknet53', pretrained=True)
m.eval()
Replace the model name with the variant you want to use, e.g. cspdarknet53
. You can find the IDs in the model summaries at the top of this page.
You can follow the timm recipe scripts for training a new model afresh.
@misc{bochkovskiy2020yolov4,
title={YOLOv4: Optimal Speed and Accuracy of Object Detection},
author={Alexey Bochkovskiy and Chien-Yao Wang and Hong-Yuan Mark Liao},
year={2020},
eprint={2004.10934},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
BENCHMARK | MODEL | METRIC NAME | METRIC VALUE | GLOBAL RANK |
---|---|---|---|---|
ImageNet | cspdarknet53 | Top 1 Accuracy | 80.05% | # 98 |
Top 5 Accuracy | 95.09% | # 98 |