Skip to content
Implementation of EfficientNet model. Keras.
Branch: master
Clone or download
sdll and qubvel Add the converter script, fixing typos (#21)
* Added the helper script to convert the weights 
* Improved README.md wording
* Fixed `inference_example.ipynb` filename typo
Latest commit 1c32ba8 Jun 18, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
efficientnet Weights for B4, B5 models (#19)(#20) Jun 17, 2019
examples Add the converter script, fixing typos (#21) Jun 18, 2019
misc Add inference example May 31, 2019
scripts Add the converter script, fixing typos (#21) Jun 18, 2019
.gitignore Create .gitignore May 31, 2019
MANIFEST.in Add setup.py May 31, 2019
README.md Add the converter script, fixing typos (#21) Jun 18, 2019
__init__.py Initial commit May 31, 2019
setup.py Updated reqs May 31, 2019

README.md

EfficientNet-Keras

This repository contains a Keras reimplementation of EfficientNet, a lightweight convolutional neural network architecture achieving the state-of-the-art accuracy with an order of magnitude fewer parameters and FLOPS, on both ImageNet and five other commonly used transfer learning datasets.

The codebase is heavily inspired by the TensorFlow implementation.

Table of Contents

  1. About EfficientNet Models
  2. Examples
  3. Models
  4. Installation
  5. Frequently Asked Questions

About EfficientNet Models

EfficientNets rely on AutoML and compound scaling to achieve superior performance without compromising resource efficiency. The AutoML Mobile framework has helped develop a mobile-size baseline network, EfficientNet-B0, which is then improved by the compound scaling method to obtain EfficientNet-B1 to B7.

EfficientNets achieve state-of-the-art accuracy on ImageNet with an order of magnitude better efficiency:

  • In high-accuracy regime, EfficientNet-B7 achieves the state-of-the-art 84.4% top-1 / 97.1% top-5 accuracy on ImageNet with 66M parameters and 37B FLOPS. At the same time, the model is 8.4x smaller and 6.1x faster on CPU inference than the former leader, Gpipe.

  • In middle-accuracy regime, EfficientNet-B1 is 7.6x smaller and 5.7x faster on CPU inference than ResNet-152, with similar ImageNet accuracy.

  • Compared to the widely used ResNet-50, EfficientNet-B4 improves the top-1 accuracy from 76.3% of ResNet-50 to 82.6% (+6.3%), under similar FLOPS constraints.

Examples

  • Initializing the model:
from efficientnet import EfficientNetB0

model = EfficientNetB0(weights='imagenet')
  • Loading the pre-trained weights:
from efficientnet import load_model

model = load_model('path/to/model.h5')

See the complete example of loading the model and making an inference in the Jupyter notebook here.

Models

The performance of each model variant using the pre-trained weights converted from checkpoints provided by the authors is as follows:

Architecture @top1* @top5* Weights
EfficientNetB0 0.7668 0.9312 +
EfficientNetB1 0.7863 0.9418 +
EfficientNetB2 0.7968 0.9475 +
EfficientNetB3 0.8083 0.9531 +
EfficientNetB4 0.8259 0.9612 +
EfficientNetB5 0.8309 0.9646 +
EfficientNetB6 - - -
EfficientNetB7 - - -

* - topK accuracy score for converted models (imagenet val set)

Installation

Requirements

  • keras >= 2.2.0 + tensorflow
  • scikit-image

Installing from the source

pip install -U git+https://github.com/qubvel/efficientnet

Installing from PyPI

pip install -U efficientnet

Frequently Asked Questions

  • How can I convert the original TensorFlow checkpoints to Keras HDF5?

Pick the target directory (like dist) and run the converter script from the repo directory as follows:

./scripts/convert_efficientnet.sh dist
  • Why are B6 and B7 model variants not yet supported?

Weights for B6-B7 have not been made available yet, but might appear soon. Follow the issue for updates.

You can’t perform that action at this time.