Nikola Bulatović - MLC 2023

Nikola Bulatović

Machine Learning Researcher, University of Montenegro, Faculty of Electrical Engineering

I am a lecturer and Machine Learning researcher.

I am currently doing my PhD research at the Faculty of Electrical Engineering, University of Montenegro (FEE-UoM), where I am working on Machine Learning algorithms for acoustic vehicle detection and speed estimation from single sensor measurements. I strongly focused on ML with Coursera’s Deep Learning Specialization course, in 2018. My professional knowledge was upgraded working as a Machine Learning research consultant in Datum Solutions where I was part of the team working on the Optical Character Recognition (OCR) system for better and faster processing and extraction of valuable information in documents.

For more than 7 years I worked as a teaching assistant at the Computer Science Department (FEE – UoM) giving lectures in Digital Image Processing, Algorithms and Programming, C, C++, and Web development. During my time at the faculty, I contributed to establishing the first BIO-ICT Center of Excellence in Montenegro where I and my team developed the SeMaR IoT system for real-time ecological monitoring.

I am one of the founders of the NGO Montenegro Robotics with the aim of educating Montenegrin youth to excel in computer science, electronics, and robotics. I developed and lectured several Arduino programming courses. I supervised the official Montenegrin team of high school students in the Robotics Olympics FIRST Global Challenge 2018, 2019, and 2020. Our team of young talents achieved notable results at the FIRST Global competition under my supervision.

During my professional career, I developed several successful and awarded software and hardware applications in the Web, Android, iOS, and IoT domains. One that stands out is the Be Responsible platform where I am one of the principal software engineers. Be Responsible is a platform enhancing two-way communication between citizens and public authorities in Montenegro, to tackle economic, ecological, traffic, and other abuses observed in everyday urban life. Be Responsible was developed in 2013 and it was awarded the winner at the competition conducted by the NGO DigitalizujMe and the United Nations Development Programme (UNDP), within the project “Open Ideas for Montenegro”. Since then, It has won several international government policies’ responsibility and transparency awards in Europe and United States.

Social Profiles

All Sessions by Nikola Bulatović

  • II DAN

    Subota, 21. maj 2022.

  • 10:00  - 10:30
    The presentation addresses novel ML methods for acoustic vehicle detection and speed estimation. Methods are based on supervised learning from single-sensor audio data and speed-dependent features extracted from data. For this research, we have collected, annotated, and published a dataset of audio-video recordings of single vehicles passing by the camera at a known constant speed. The dataset contains 304 urban-environment real-field recordings of ten different vehicles which is the most extensive annotated dataset for vehicle speed estimation. The methods are able to accurately predict the pass-by instant of a vehicle and to estimate its speed with an average error of 6.92 km/h.
    Where
    Velika Terasa, Palić




Share

Nikola Bulatović