Interested in Better SIFT, Revise Text & Digit Recognition and Implementation of QR Codes

[Muzamil Sethar]

Hello everyone. I am a student of computer science, currently in the 6th semester at the National University of Computer and Emerging Sciences Islamabad. I am interested to work with OpenCV because of my interest in image processing and machine learning. I want to build/improve things using my skills and hard work. I am proficient in C++ and my knowledge in image processing is 8/10 at the student level (Not proficient, but I know basics). Currently, I am taking a machine learning course and I like it very much. But due to COVID-19, our classes are discontinued but I am still working from home and exploring algorithms.

I was interested in improved RANSAC but in the mailing list, I found a topic in which it was said that RANSAC might not be the best choice, so I had to switch to other choices which include "Better SIFT in the main repository", "Revise/improve Text & Digit Recognition Samples" and "Implementation of QR-codes decoder and encoder". To be honest, I am new to open source and I used not to share my code with anyone because of the mentality and fear of not sharing code with anyone due to the warnings by the college about plagiarism cases. But a few days ago, I uploaded all my work which can help others and I found it very delightful. Here's my GitHub account link "https://github.com/Muzamil-Hussain".

I need help to choose the best idea to work on and explore more and apply for it in GSOC 2020. I might not know all the things, but I am interested to learn and explore more and contribute.

Thank you.

[Vitaly Tuzov]

Hi,

You have to decide on the project you are going to do through this GSoC and prepare your proposal for it. Projects you mention relates to different areas of computer vision and require different skills and activities during this summer. SIFT is classic computer vision project that mainly require performance tuning. T&D recognition samples are Deep NNet project that require working on UX, code cleanness and readability. QR-codes is also classic computer vision that at the moment require working on corner cases and improvement of detection reliability for low quality and somehow distorted images AFAIK. So you have to choose the one which fascinate you most of all.

After that you have to submit your application for GSoC2020 via summerofcode.withgoogle.com till March 31th.

Best regards,

Vitaly

среда, 25 марта 2020 г., 14:03:37 UTC+3 пользователь Muzamil Sethar написал: