Subscribe

UiPath AI Center

UiPath AI Center

Image Classification

Out of the Box Packages > UiPath Image Analysis > Image Classification

This preview model is a retrainable deep learning model used to classify images. You can train it on your own data and create an ML Skill to perform image classification. This ML Package must be retrained, if deployed without training first, deployment will fail with an error stating that the model is not trained.

Model details

Input type

FILE

Input description

Full path of the image file on which you want to classify.
Make sure that the image format is either JPEG or PNG.

Output description

JSON with identified label for the image and confidence score (between 0-1).

{
  "response": {
    "label": "car",
    "confidence": 0.85657345056533813
  }
}

Recommend GPU

By default, a GPU is recommended.

Training enabled

By default, training is enabled.

Pipelines

All three types of pipelines (Full Training, Training, and Evaluation) are supported by this package. For most use cases, no parameters need to be specified, the model is using advanced techniques to find a performant model. In subsequent trainings after the first, the model uses incremental learning (that is, the previously trained version will be used, at the end of a Training Run).

Dataset format

For training and evaluation datasets, point to a folder with a subfolder called Images and this subfolder can contain input several folders with different classes (for example, a folder called cats with pictures of cats, and another one called dogs with pictures of dogs, and so on).
Example :

-- <Training / Evaluation Directory>
   -- images
      -- Bus
         -- bus001.jpg
         -- bus002.jpg
         -- bus003.jpg
      -- Truck
         -- truck001.jpg
         -- truck012.png
         -- truck0030.jpeg
      -- Car

CSV file format:
Each CSV file can have any number of columns, but only two will be used by the model. Those columns are specified by the parameters input_column (if not modified, the default value is input) and target_column (if not modified, the default value is target).

Environment variables

  • Epochs - default value 20

Artifacts

Classification report

precision    recall  f1-score   support

    Positive       0.75      0.90      0.82        10
    Negative       0.88      0.70      0.78        10

    accuracy                           0.80        20
   macro avg       0.81      0.80      0.80        20
weighted avg       0.81      0.80      0.80        20

Confusion matrix

Predictions.csv

This is a CSV file with predictions on the test set used for evaluation.

filename      actual        predicted
38  00043.jpg   Positive    Positive
17  00001.jpg   Positive    Positive
59  00014.jpg   Negative    Positive
31  00015.jpg   Positive    Positive
15  00008.jpg   Positive    Positive
69  00025.jpg   Negative    Negative
49  00003.jpg   Positive    Positive
5     00034.jpg Positive    Positive
36  00044.jpg   Positive    Positive
50  00042.jpg   Negative    Positive
96  00011.jpg   Negative    Negative
53  00046.jpg   Negative    Positive
94  00036.jpg   Negative    Negative

Updated about a month ago

Image Classification


Suggested Edits are limited on API Reference Pages

You can only suggest edits to Markdown body content, but not to the API spec.