| Training Techniques | RMSProp, Weight Decay, Label Smoothing, AutoAugment, Stochastic Depth |
|---|---|
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
| ID | tf_efficientnet_b0 |
| SHOW MORE |
| Training Techniques | RMSProp, Weight Decay, Label Smoothing, AutoAugment, Stochastic Depth |
|---|---|
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
| ID | tf_efficientnet_b1 |
| SHOW MORE |
| Training Techniques | RMSProp, Weight Decay, Label Smoothing, AutoAugment, Stochastic Depth |
|---|---|
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
| ID | tf_efficientnet_b2 |
| SHOW MORE |
| Training Techniques | RMSProp, Weight Decay, Label Smoothing, AutoAugment, Stochastic Depth |
|---|---|
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
| ID | tf_efficientnet_b3 |
| SHOW MORE |
| Training Techniques | RMSProp, Weight Decay, Label Smoothing, AutoAugment, Stochastic Depth |
|---|---|
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
| ID | tf_efficientnet_b4 |
| SHOW MORE |
| Training Techniques | RMSProp, Weight Decay, Label Smoothing, AutoAugment, Stochastic Depth |
|---|---|
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
| ID | tf_efficientnet_b5 |
| SHOW MORE |
| Training Techniques | RMSProp, Weight Decay, Label Smoothing, AutoAugment, Stochastic Depth |
|---|---|
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
| ID | tf_efficientnet_b6 |
| SHOW MORE |
| Training Techniques | RMSProp, Weight Decay, Label Smoothing, AutoAugment, Stochastic Depth |
|---|---|
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
| ID | tf_efficientnet_b7 |
| SHOW MORE |
| Training Techniques | RMSProp, Weight Decay, Label Smoothing, AutoAugment, Stochastic Depth |
|---|---|
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
| ID | tf_efficientnet_b8 |
| SHOW MORE |
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
|---|---|
| ID | tf_efficientnet_el |
| SHOW MORE |
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
|---|---|
| ID | tf_efficientnet_em |
| SHOW MORE |
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
|---|---|
| ID | tf_efficientnet_es |
| SHOW MORE |
| Training Techniques | Noisy Student, FixRes, RMSProp, Weight Decay, Label Smoothing, AutoAugment, RandAugment |
|---|---|
| Architecture | 1x1 Convolution, Average Pooling, Convolution, Dense Connections, Dropout, Inverted Residual Block, Batch Normalization, Squeeze-and-Excitation Block, Swish |
| ID | tf_efficientnet_l2_ns_475 |
| SHOW MORE |
EfficientNet is a convolutional neural network architecture and scaling method that uniformly scales all dimensions of depth/width/resolution using a compound coefficient. Unlike conventional practice that arbitrary scales these factors, the EfficientNet scaling method uniformly scales network width, depth, and resolution with a set of fixed scaling coefficients. For example, if we want to use $2^N$ times more computational resources, then we can simply increase the network depth by $\alpha ^ N$, width by $\beta ^ N$, and image size by $\gamma ^ N$, where $\alpha, \beta, \gamma$ are constant coefficients determined by a small grid search on the original small model. EfficientNet uses a compound coefficient $\phi$ to uniformly scales network width, depth, and resolution in a principled way.
The compound scaling method is justified by the intuition that if the input image is bigger, then the network needs more layers to increase the receptive field and more channels to capture more fine-grained patterns on the bigger image.
The base EfficientNet-B0 network is based on the inverted bottleneck residual blocks of MobileNetV2, in addition to squeeze-and-excitation blocks.
To load a pretrained model:
import timm
m = timm.create_model('tf_efficientnet_b0', pretrained=True)
m.eval()
Replace the model name with the variant you want to use, e.g. tf_efficientnet_b0. 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{tan2020efficientnet,
title={EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks},
author={Mingxing Tan and Quoc V. Le},
year={2020},
eprint={1905.11946},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
Image Classification on ImageNet
| MODEL | TOP 1 ACCURACY | TOP 5 ACCURACY |
|---|---|---|
| tf_efficientnet_l2_ns_475 | 88.24% | 98.55% |
| tf_efficientnet_b8 | 85.35% | 97.39% |
| tf_efficientnet_b7 | 84.93% | 97.2% |
| tf_efficientnet_b6 | 84.11% | 96.89% |
| tf_efficientnet_b5 | 83.81% | 96.75% |
| tf_efficientnet_b4 | 83.03% | 96.3% |
| tf_efficientnet_b3 | 81.65% | 95.72% |
| tf_efficientnet_el | 80.45% | 95.17% |
| tf_efficientnet_b2 | 80.07% | 94.9% |
| tf_efficientnet_b1 | 78.84% | 94.2% |
| tf_efficientnet_em | 78.71% | 94.33% |
| tf_efficientnet_es | 77.28% | 93.6% |
| tf_efficientnet_b0 | 76.85% | 93.23% |
