metadata

license: apache-2.0
tags:
  - RyzenAI
  - image-classification
  - onnx
datasets:
  - imagenet-1k

EfficientNet

The EfficientNet model was proposed in EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks by Mingxing Tan and Quoc V. Le. EfficientNets are a family of image classification models, which achieve state-of-the-art accuracy, yet being an order-of-magnitude smaller and faster than previous models. The specific version of EfficientNet here is EfficientNet-ES (EdgeTPU-Small).

We develop a modified version that could be supported by AMD Ryzen AI.

Model description

The abstract from the paper is the following:

Convolutional Neural Networks (ConvNets) are commonly developed at a fixed resource budget, and then scaled up for better accuracy if more resources are available. In this paper, we systematically study model scaling and identify that carefully balancing network depth, width, and resolution can lead to better performance. Based on this observation, we propose a new scaling method that uniformly scales all dimensions of depth/width/resolution using a simple yet highly effective compound coefficient. We demonstrate the effectiveness of this method on scaling up MobileNets and ResNet. To go even further, we use neural architecture search to design a new baseline network and scale it up to obtain a family of models, called EfficientNets, which achieve much better accuracy and efficiency than previous ConvNets. In particular, our EfficientNet-B7 achieves state-of-the-art 84.3% top-1 accuracy on ImageNet, while being 8.4x smaller and 6.1x faster on inference than the best existing ConvNet. Our EfficientNets also transfer well and achieve state-of-the-art accuracy on CIFAR-100 (91.7%), Flowers (98.8%), and 3 other transfer learning datasets, with an order of magnitude fewer parameters.

The original code can be found here.

Intended uses & limitations

You can use the raw model for image classification. See the model hub to look for fine-tuned versions on a task that interests you.

How to use

Installation

Follow Ryzen AI Installation to prepare the environment for Ryzen AI.
Run the following script to install pre-requisites for this model.
```
pip install -r requirements.txt 
```

Test & Evaluation

Inference one image (Image Classification):

import onnxruntime
import argparse
from PIL import Image 
import torchvision.transforms as transforms 

parser = argparse.ArgumentParser()
parser.add_argument('--onnx_path', type=str, default="EfficientNet_int.onnx", required=False)
parser.add_argument('--image_path', type=str, required=True)

args = parser.parse_args()

def read_image():
  # Read a PIL image 
  image = Image.open(args.image_path)
  normalize = transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
  
  transform = transforms.Compose([
    transforms.ToTensor(),
    transforms.Resize((224, 224)),
    normalize,
  ])
  img_tensor = transform(image).unsqueeze(0)
  return img_tensor.numpy()


def main():
  so = onnxruntime.SessionOptions()
  ort_session = onnxruntime.InferenceSession(
    args.onnx_path, so, providers=['CUDAExecutionProvider'])
  ort_inputs = {
    "WrapModel::input_0": read_image()
  }
  output = ort_session.run(None, ort_inputs)[0]
  print("class id =", output[0].argmax())


if __name__ == "__main__":
  main()

Evaluate ImageNet validation dataset(50,000 images), using eval_onnx.py .

python eval_onnx.py --onnx_model EfficientNet_int.onnx --ipu --provider_config Path\To\vaip_config.json --data_dir /Path/To/Your/Dataset

Performance

Dataset: ImageNet validation dataset (50,000 images).

Metric	Accuracy on IPU
top1& top5 accuracy	77.72% / 93.78%

Citation

@article{EfficientNet,
 author       = {Mingxing Tan and Quoc V. Le},
  title       = {Searching for MobileNetV3},
  year        = {2019},
  url         = {https://arxiv.org/abs/1905.11946},
}