The learner is expected to demonstrate the capability of the project by consolidating all the relevant modules pertaining to the Project.
Prerequisites and Knowledge/Skills needed to undertake this Project Basic Knowledge of Programming Languages like Java, Python, R Programming and SQL Before Commencing the project the following links have to be examined.
https://www.w3schools.com/
https://www.edx.org/
https://www.udacity.com/
Data scientist
Project Background and Description Handwriting to text converter is a Machine learning model that can scan the handwritings from a paper and convert it into text format. The algorithm will be trained in a phase wise manner where it will be taught to recognize various letters of English alphabets in various forms (cursive, italics, monospaced).The next phase will teach the algorithm to identify groups of letters/words. The final phase will teach the algorithm to recognize groups of words/sentences
Project Scope The project has high scope in the market since nowadays every written document is getting converted into E-Documents. It is easy for students to make notes and share it within their groups or to their teachers. Helps in reading files without any problem of bad handwriting or mispronunciation.
High-Level Requirements Person should have extended knowledge in Machine Learning, convolutional neural networks, recurrent neural networks etc.
Deliverables The learner is expected to demonstrate the capability of the project by consolidating all the relevant modules pertaining to the Project.
Material or Infrastructure Requirements A fully functional PC/Laptop with necessary software installed for this project, downloaded from Open Source.
Camera module with decent resolution Dataset required to train the project.
Specific Exclusions from Scope This project has a constraint that it can still not recognize some characters that are hard to recognize.
Downloading and Installing Python from open source
Importing all necessary libraries and modules for machine learning
Install Tensorflow or torch to run and simulate CNN.
High-Level Timeline/Schedule Week 1: Introduction to backbone of the ML model and its features Week2: Collect/create necessary dataset for the model Week 3: Brief about the usage of various libraries Week 4: To help creating a basic framework for the model.
Week 5: Using different dataset, let the model learn how to recognize different writings Week 6: Test the model with new dataset and check it’s accuracy Week 7: Improve the model’s accuracy with more datasets Week 8: Get final results of the model Week 9: Full-fledged project development to be completed
Assumptions The learner is required to work on the project from the beginning based on the input and the allocation that is made every week.
Guidelines to be formed and followed to maintain uniform coding ethics and documentation style
Constraints There should be no violation of rules or law. Project should be free from plagiarism.
Risks There should be no inaccuracy in conversion of handwriting; it may lead to change in meaning of the sentence or lead to mistakes in the document.
