Image Embedding with data2vec

author: David Wang

Description

This operator extracts features for image with data2vec. The core idea is to predict latent representations of the full input data based on a masked view of the input in a self-distillation setup using a standard Transformer architecture.

Code Example

Load an image from path './towhee.jpg' to generate an image embedding.

Write a pipeline with explicit inputs/outputs name specifications:

from towhee import pipe, ops, DataCollection

p = (
    pipe.input('path')
        .map('path', 'img', ops.image_decode())
        .map('img', 'vec', ops.image_embedding.data2vec(model_name='facebook/data2vec-vision-base-ft1k'))
        .output('img', 'vec')
)

DataCollection(p('towhee.jpeg')).show()

Factory Constructor

Create the operator via the following factory method

data2vec(model_name='facebook/data2vec-vision-base')

Parameters:

​ model_name: str

The model name in string. The default value is "facebook/data2vec-vision-base-ft1k".

Supported model name:

• facebook/data2vec-vision-base-ft1k
• facebook/data2vec-vision-large-ft1k

Interface

An image embedding operator takes a towhee image as input. It uses the pre-trained model specified by model name to generate an image embedding in ndarray.

Parameters:

​ img: towhee.types.Image (a sub-class of numpy.ndarray)

​ The decoded image data in towhee.types.Image (numpy.ndarray).

Returns: numpy.ndarray

​ The image embedding extracted by model.