Pushkar.

Hello, I’m

Pushkar

Deep Learning Researcher

Hero
Instagram Twitter Github LinkedIn SoundCloud

ABOUT ME

Hi There! I'm Pushkar Jajoria

Deep Learning Researcher

Hi, I'm Pushkar I am a Deep Learning researcher. I am currently working for the University of Galway on the Polifonia project.

My aim in life is to advance and use the methods of Deep learning to solve interesting and challenging problems. My current research is on Audio-Lyrics Alignment for Italian Arias using transfer learning. To this end, I create DL methods for modeling both symbolic and audio music which allows music composers, musicians, and musicologists to leverage these models.

I offer consultations for companies in picking the right and the best solution for their use case. Feel free to drop me a message to get in touch.

I like playing my guitar (Ibanez RG370) and composing music. I try to use machine learning models to help me compose better music.

I'll be happy to meet and discuss new ideas, and opportunities, or just have a casual chat over coffee.

Download CV

PROJECTS

Sequential Prediction of Polyphonic Classical Music

This dissertation tackles the prediction of polyphonic music using a human-interpretable model, presenting an alternative to existing "blackbox" deep learning methods. By implementing a nested Markov process and defining interpretable musical features, the study effectively navigates the vast prediction space and enhances the understanding of musical styles. Trained on J.S. Bach Chorales and Mozart Sonatas, the model achieves impressive prediction accuracy while demonstrating an understanding of nuanced musical rules.

Robust Journey Planning

With a focus on enhancing user experience and providing accurate results, our collaborative project delivers a Robust Journey Planning system that takes travel delays into account, calculates optimal routes, and employs advanced algorithms and machine learning techniques to improve the probability of successful trips. The project's well-organized structure, encompassing various notebooks for data preparation, graph creation, delay prediction, route generation, and validation, ensures a robust and efficient travel planning solution for users in Zurich, Switzerland.

Towards a computational definition of Tresillo rhythm and its tracing in pop music.

This project investigates the use and popularity of the Tresillo rhythm in popular music. It defines and formalizes the rhythm, then traces its presence in the US Billboard Top 20 Charts over the past 20 years. Through similarity calculations, it measures how closely a song's rhythm aligns with the Tresillo pattern. The study utilizes different formalizations and comparison methods for validation. The findings shed light on the Tresillo rhythm's prevalence and offer insights into its usage in popular music.

CONTACT ME

Contact Info

Let's connect! I'm always keen to communicate and collaborate with like-minded individuals and organizations. Whether you've got an inquiry, an exciting opportunity, or simply want to share your thoughts, feel free to reach out.

Email

FirstnameLastname@gmail.com
Visit my social profile and get connected
Instagram Twitter Github LinkedIn SoundCloud

This free static hosting site is powered by Porkbun.com.