Task 1: The goal of the task is to collect handwritten samples, so we can train machine learning models to automatically identify what people are writing. We need to collect two kinds of samples:
text annotation like scribbling or circling: we call these "gestures".
Please note that all writing will be done with a stylus on a tablet screen; you cannot use your finger to write on the screen for these tasks. Also please note that the app is intended to be used full-screen.
Task 2: The purpose of this task is to collect digital handwritten text samples, composed of multiple lines and annotate them. We will use those samples to train machine learning models to localize text.
To do this, we need examples of writing with the indication of word and line separation, from which the computer can learn how to do it automatically. You will be shown the text the model is expected to answer, and will be asked to handwrite it on the device, with the same line breaks.
Task 3: The purpose of this task is to collect digital handwritten text samples, composed of a single line. We will use those samples to train machine learning models to read text.
- Quality Expectation - more than 98%
The interface of the tool:
- Numbers/annotations will be actively tracked
- Total time spent on the tool will be captured
- Tech Mahindra NDA Signed (This will be issued later in the project)
- Passed Initial annotation Review
- Data collection consent form to be signed/ acknowledged on the app
Pay and Duration : The duration of this task is 6-7 hours, and the task should be completed in the same day and the payment will be made through freelacer only.
for 1 hour a freelancer will be given 15$