The following table lists the automated processors currently available in Taskmonk.

• The first column lists the processor type, organizing the available processors into pre-processing and post-processing tools.

• The second column lists the processor subtype. This indicates when they can be used.

• The third column lists the processor’s area of function.

• The fourth column lists the name of the processor. Click the name of the processor to view details on using it.

PDF Data Extraction

Open

To configure data extraction from input PDF files:

1. Select field to extract data from PDF: Click to select a field that exists in the PDF file and contains the data that you want to use for labeling.

2. Select field to store the output: Click to select the field into which the PDF Data Extraction tool must write the data extracted from the data field in the PDF file.

3. Click Add to include this capability into your project.

Named Entity Recognition

Open

To configure the Named Entity Recognition tool:

1. Click to Select the input field with text to extract entities. This is the field that contains the entities that you want to include into your labeling.

2. Click to Select the provider of named entity recognition.

3. Click to Select the output field to store the extracted entities.

4. Click Add to include this capability into your project.

Grammar Check

The Grammar Check tool checks the spelling of words in fields to which it is directed.

To configure the Grammar Check tool:

1. Click to Select the field on which to run spell-check.

2. Click to Select field to store spell check output.

3. Click Add to include this capability into your project.

Web Scraping

Use the Web Scraping tool to extract data from online sources.

To configure the Web Scraping tool:

1. Click to Select the field with the scrape URL.

2. Click to Select the field to store the output.

3. Click Add to include this capability into your project.

Text Translation

Use the Text Translation tool to automatically translate input text into the target language.

1. Click to Select the Translation Provider that you want to use.

2. Click to Select the Field to Store the Source Language.

3. Click to Select the input language code. This is the language in which the source data is expected to be coded.

4. Click to Select the language code in which the input needs to be converted.

5. Click to Select the field to store the translated text.

6. Click Add to include this capability into your project.

Text Transliteration

Use the Text Transliteration tool to configure how you want to transliterate text from one language to another.

1. Click to Select the Transliteration Provider.

2. Click to Select the field that contains the source text. This is the text that must be transliterated.

3. Click to Select the source language. This is the language of the source text.

4. Click to Select the source script to input text. This is the script in which the input text is written. Thus, for example, you could enter text in Hindi written in English script.

5. Click to Select the target script into which the input text will be converted.

6. Click to Select the field that must store the transliteration output.

7. Click Add to include this capability into your project.

Phrase Extractor

Use the Phrase Extractor tool to identify and extract phrases that have a strong meaning.

1. Click to Select the input field. This is the field that contains the untagged text.

2. Enter the Anchor Word that must be used to identify phrases. For example, Big Data.

3. Enter the Prefix Count to specify the number of words preceding the Anchor Word that may be a part of the phrase.

4. Enter the Suffix Count to specify the number of words following the Anchor Word that may be a part of the phrase.

5. Click POS to specify the part of speech that the phrase should be tagged with.

6. Click Extractor Code Language to specify the language in which the extractor code is provided.
   Supported languages are:

   1. Python.

   2. Java.

7. Enter the Extraction Code in the language specified.

8. Click Add to include this capability into your project.

NLP

Use the NLP tool to tag and extract entities from text.

1. Click to Select the input field. This is the field that contains the untagged text.

2. Click to Select the annotations field. This is the field that will contain the automatically tagged text.

3. Click to specify whether you want to Filter common nouns.

4. Enter the Comma-separated list of entities that you want to tag in the input text.

5. Click to Select the provider of the NLP service.

6. Click Add to include this capability into your project.

Field Dictionary Matcher

Use the Field Dictionary Matcher to connect to your dictionary in Taskmonk and provide settings associated with its input, output, and settings.

1. Click to Select the input field that contains the text that you want to run through the dictionary.

2. Click to Select the output field that must contain the dictionary’s output.

3. Enter the Key for the Dictionary. This key enables you to access your dictionary. Once your dictionary is created, you should receive this key from Taskmonk.

4. Click Case Sensitive Match and specify whether you want the matching to be case sensitive.

5. Click Add to include this capability into your project.

Extract Audio

Use the Extract Audio tool to extract audio content from video files.

1. Click to Select the field that contains the view URL.

2. Click to Select the field that must store the extracted audio.

3. Click Add to include this capability into your project.

Audio Transcription

Use the Audio Transcription tool to automatically create transcription from audio files.

1. Click to Select the field that stores the URL to the input audio file.

2. Click to Select the provider of the audio transcription.

3. Click to Select the field that must store the extracted text results.

4. Click Add to include this capability into your project.

Split Audio On Pause

Use the Split Audio On Pause tool to automatically split input audio into multiple files on detecting a pause of a given duration.

1. Click to Select the field containing the input URL.

2. Click to Select the field to store the output after the audio is split.

3. Enter the Duration to split in seconds as the minimum duration of the pause at which the audio should be split.

4. Click Add to include this capability into your project.

Split Audio in Separate Channels

Use the Split Audio in Separate Channels tool to automatically split input audio into multiple files based on the number of channels present.

1. Click to Select the field containing the input audio URL.

2. Click to Select the field to store the output after the audio is split.

3. Click Add to include this capability into your project.

Video Type Convertor

Use the Video type Convertor tool to convert video files of different formats into a format that Taskmonk uses.

1. Click to Select the field that contains the video URL.

2. Click to Select the field that must store the processed video file.

3. Click Add to include this capability into your project.

Load Annotations

Use the Load Annotations tool to load annotation data (that you already have) from a separate file on to another image. This will simplify labelers' tasks, requiring them to only move the annotation edges around until they accurately surround the object concerned.

1. Click to Select the data format of the annotation source file.

2. Click to Select the data format to which the source annotation must be converted.

3. Click to Select the field that contains the source annotations.

4. Click to Select the field that must store the converted annotations.

5. Click Add to include this capability into your project.

Image Details Extraction

Use the Image Details Extraction field to store image-related details from input images.

1. Click to Select the field that contains the source image path.

2. Click to Select the field that must store the image’s height.

3. Click to Select the field that must store the image’s width.

4. Click to Select the field that must store the RGB value of the background color. This is an optional field.

5. Click to Select the field that must store the object-to-background pixel ratio.

6. Click Add to include this capability into your project.

Object Detection

Use the Object Detection tool to automatically detect objects in images when they load on to the labeling UI.

1. Click to Select the field that contains URLs to the images that must be labeled.

2. Click to Select the object detection service provider.

3. Click to select the field that must store the generated annotations.

4. Use the Set External URL field to provide the URL to the custom object detection service provider. If you selected Google as the service provider, you do not need to provide any detail here.

5. Click Add to include this capability into your project.

Color Extraction

Use the Color Extraction tool to extract primary and secondary color information from images..

1. Click to Select the input field that contains the image URL.

2. Click to Select the provider of the color extraction software.

3. Click to Select the output field that must store the extracted color information.

4. Click Add to include this capability into your project.

TifftoPNG Convertor

Use the TIFF-to-PNG convertor to convert images in TIFF format to other image formats.

1. Click to Select the output format to which the image file should be converted.

2. Click to Select input field into save the original URL.

3. Click Add to include this capability into your project.

Pose Detector

Use the Pose Detector to detect human beings in images and identify their poses.

1. Click to Select the input field that contains the image that must be processed.

2. Click to Select the field containing the annotations.

3. Click Add to include this capability into your project.

Keypoint Detector

Use the Keypoint Detector tool to detect keypoints or interest points in a person identified in an image or a video.

1. Click to Select the input field that contains the image or the video that must be processed.

2. Click to Select the annotation field to which the output must be saved.

3. Click to Select the Type of Detector that you wish to employ. These include:

   1. 18 point pose keypoints.

   2. 21 point hand keypoints.

   3. 68 point face keypoints.

   4. 468 point face mesh.

4. Enter the Detection FPS to be used for detection

5. Click Add to include this capability into your project.

Search LinkedIn

Use the Search LinkedIn tool to automatically search LinkedIn for details associated with specific people and organizations.

1. Click to Select the field containing the input URL. This is the field that contains a document listing out the people and organizations that you want to search on LinkedIn.

2. Click to Select the field to store the search result.

3. Click Add to include this capability into your project.

Merge Fields

Use the Merge Fields tool to merge the contents of two or more fields into a single string.

1. Click to Select the fields that you want to merge.

2. Click to Select the field into which the merged output must be stored.

3. Click Add to include this capability into your project.

File Download

Use the File Download tool to automatically download a file on task creation.

1. Click to Select the field with the URL of the file to be downloaded.

2. Click Add to include this capability into your project.

Code Runner

Use the Code Runner tool to run a set code when a specific event occurs. This could be on task creation, on level change, and so on.

1. Click to Select the code language that you want to use and enter the Code to be run in the field provided.

2. Click Add to include this capability into your project.

TmAnnotations Data Convertor

Use the TmAnnotations Data Convertor tool to convert annotation data generated by Taskmonk to the format suitable for your downstream application.

1. Click to Select the field containing the annotations to be converted.

2. Click to Select the field that has the input URL for which the annotations are generated.

3. Enter the Code to be run to generate the output annotation file,

4. Specify the folder name present in your bucket to which you want to upload the annotations.

5. Click to Select the field where the URL of the output file is to be saved.

6. Click Add to include this capability into your project.

Optical Character Recognition

Use the Optical Character Recognition to automatically read text in input images.

1. Click to Select the input field containing the path to the image.

2. Click to Select the OCR service provider from the drop-down menu.

3. Click Add to include this capability into your project.

Output Formatter

Use the Output Formatter tool to convert annotations created in Taskmonk to another format. Click to Specify the output format and click Add to include this capability into your project.