Efficientnet Embedding Operator

Authors: kyle he

Overview

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[1], which is trained on imagenet dataset.

Interface

__init__(self, model_name: str = 'efficientnet-b7', framework: str = 'pytorch', weights_path: str = None)

Args:

• model_name:
  • the model name for embedding
  • supported types: str, for example 'efficientnet-b7'
• framework:
  • the framework of the model
  • supported types: str, default is 'pytorch'
• weights_path:
  • the weights path
  • supported types: str, default is None, using pretrained weights

__call__(self,  image: 'towhee.types.Image')

Args:

• image:
  • the input image
  • supported types: towhee.types.Image

Returns:

The Operator returns a tuple Tuple[('feature_vector', numpy.ndarray)] containing following fields:

• feature_vector:
  • the embedding of the image
  • data type: numpy.ndarray
  • shape: (dim,)

Requirements

You can get the required python package by requirements.txt.

How it works

The towhee/efficientnet-image-embedding Operator implements the function of image embedding, which can add to the pipeline. For example, it's the key Operator named embedding_model within image-embedding-efficientnetb5 and image-embedding-efficientnetb7 pipeline.

Reference

[1].https://github.com/lukemelas/EfficientNet-PyTorch#example-feature-extraction