Master Thesis in Machine Learning

Here is an exciting Master Thesis opportunity in Machine Learning: Automated SQL Command Generation from Digital Diagrams Using Computer Vision Techniques.

About this thesis opportunity

This master thesis project will be hosted by Consafe Logistics in Lund and will be part of a team of extraordinary developers with different backgrounds, experiences, with specializations in Machine Learning. You will join and be part of a professional environment, excellent for newly graduated engineers in terms of learning and networking.

Objective:
The goal of this thesis is to develop a computer vision-based system that interprets digital CAD drawings and translates them into SQL commands. This approach would save users a lot of time in the implementation phase for one of our AI-modules in Astro WMS. The goal is to automatically generate corresponding SQL code to create and populate database tables.

Methodology:

1. Image Preprocessing and Detection: Develop preprocessing algorithms to enhance images, focusing on detecting shapes, lines, and text in the diagrams.

1. Object Recognition and Classification: Use deep learning models (e.g., convolutional neural networks) to recognize elements in a warehouse CAD diagram

1. Text Recognition and Parsing: Apply OCR (Optical Character Recognition) to identify and read text labels within the diagram, classifying them as table names, column names, or constraints.

1. SQL Generation Module: Develop a module that translates recognized elements and relationships into SQL statements (e.g., CREATE TABLE, INSERT INTO) using natural language processing (NLP) and rule-based logic.

Expected Outcomes:
The system will automatically interpret key elements of digitally drawn database schema and output corresponding SQL commands to create and populate tables, reducing the need for manual SQL coding and making database design more accessible for one of our AI-modules.

What you will bring

We are looking for a student (or pair of students) who is knowledgeable, creative and an avid problem-solver. You should be studying a master program in computer science, machine learning or equivalent at a Swedish university or through a valid exchange program. The student must have a solid understanding of machine learning methodologies and programming skills; Python, NumPy, Anaconda, TensorFlow to name a few. Good understanding of C/C#, SQL will be considered a merit.

Why join Consafe Logistics?

Consafe Logistics is a software development company. We allow developers to tackle real-world challenges, optimizing systems and processes that directly impact efficiency and delivery of warehouses. With rapid advancements in AI and automation, Consafe Logistics offers a dynamic field with plenty of opportunities for innovation and growth.

What happens after you apply?

Applications should include a brief personal letter, and a CV. Experience deemed relevant is highly valued. Describe the project and your role in it. Suitable applicants will be interviewed as soon as applications are received.

Contact for this thesis opportunity: Eltayeb.Bayomi@Consafelogistics.com

Ansök här