To Fellow Readers:
Thanks for stopping by my blog, where I will be (in the coming weeks) updating you on my journey of completing my Senior Project. I am really excited to finally be starting work on this project, one that I have been looking forward since the start of the school year.
My project is officially known as “MUSIC SELECTION BASED ON PHOTO AND SCENE ANALYSIS WITH MACHINE LEARNING”, or MSBOPASAWML for short. It seems like a daunting name, though I can assure you its not, as I came up with it :-P. Before I explain what the project is about, I want to give you a bit of background on me and why I chose to do this project.
I have been playing music since first grade, which is when I started learning piano. In fourth and fifth grade, my elementary school had a band and orchestra program, and I picked up the saxophone. Continuing in middle school, I played the saxophone in our school’s 7th and 8th grade orchestra. I continued piano into high school, and took piano class at school in 11th grade.
As for my computer science background, I have been coding since sixth grade, where I made my first ever website, which was an online version of the Hitchhiker’s Guide to the Galaxy. Since then, I have worked on many programming projects, participated in many coding competitions and hackathons, hosted a competitions and a hackathon, and currently am a member of the BISV Competitive Computer Science Club. In my high school years, I had mostly been doing procedural programming and creating applications. As you can tell, I have experience in the fields of CS and Music, so I wanted to take on a project that took on a bit of both, as I really enjoy both subjects.
Back to my project’s “MUSIC SELECTION BASED ON PHOTO AND SCENE ANALYSIS WITH MACHINE LEARNING”. The basic idea behind it is, given a set of photos from the user, the program should output optimal background music options based on what it sees in those photos. I had this idea as everytime I am tasked with creating slideshows of photos, I take a very long time trying to find the right music with the right mood to match the photos. There are some software that can do similar tasks, though they are all rudimentary and have a very limited song set. I hope that by completing this project I can not only create this program that can help me and others in creating photo slideshows, but also be able to discover even more about music and machine learning, as the concept of music relies much more upon feelings and emotions rather than mathematics and numbers like computers use.
This week, I have just been taking lessons from my Machine Learning course, which I am learning a lot about the math behind ML as well as some concepts of Machine Learning code. I don’t have much to report about it (though it is very fun seeing some of the concepts that I have learned about in Math or read about and physically implementing it. I will be going much more in depth about my physical code once I finished learning about the code.
Thank you so much for reading my blog post for this week. It is a bit longer (as I was giving more exposition to the project), expect shorter blog posts in weeks 2 and 3, where I will mostly be working on learning about the code and topic. I hope you enjoyed my blog post, and hope to see you again next week!
-Ray Lin ._.