Object Embedding

Model by Modzy

This model produces embeddings from images. The model accepts JPEG images of any size. As output, it produces a representative image embedding in a JSON format.

This model can be used in multiple ways. It can be used as a step in determining image similarity /dissimilarity, image clustering, or scene recognition.

  • Description

    Product Description

    PERFORMANCE METRICS

    Adversarial Defense – This model has a built-in adversarial defense feature. Click here to read more about Modzy’s proprietary adversarial defense strategy.

    40% Bleu – A method for assessing the quality of text that has been machine-translated from one language to another. The closer the machine translation is to expert human translation, the better the score. Further information here.

    This model was trained by Facebook on the WMT-19 dataset. It achieves a Bleu score of 40.0.

    OVERVIEW:

    This model uses the Facebook Fairseq sequence modeling toolkit, which is PyTorch based.

    TRAINING:

    No information available.

    VALIDATION:

    No information available.

    INPUT SPECIFICATION

    The input(s) to this model must adhere to the following specifications:

    Filename Maximum Size Accepted Format(s)
    image 1M .jpg, .png, .tif

    OUTPUT DETAILS

    This model will output the following:

    Filename Maximum Size Format
    results.json 1M .json