I am Goce Dimitrov
Name: Goce Dimitrov
Profile: Computer Science & Electronics MEng
Email: contact@gocedimitrov.com
Resume: Download
Skills
I am a passionate Software Engineer from Kochani, Macedonia, specialising in building and designing exceptional websites, applications and anything else in between.
Graduated from the University of Bristol, UK in July, 2019. After graduation did a 3 month internship at the Bristol Vision Institute, where I developed a software for Subjective and Objective video evaluation. Currently I am a Freelancer working on design and development of websites, but ready to accept a full time position in a Company at any time.
I love challenging myself into learning and doing interesting things that matter. My abundant energy fuels me in the pursuit of many interests, hobbies and areas of study. I’m a fast learner, able to pick up new skills and juggle different projects and roles with relative ease.
Online Assessment Platform for Online assessment of Vocational skills for a Workers UniversityFULL-STACKWEBSITEOnline Assessment PlatformA responsive and intuitive full-stack website designed and developed using various modern and secure techologies for a Vocational University that allows Online Assessment of candidates. The website allows candidates to choose their desired profession, apply and submit their relevant information and documents, pay for the assessment and take a online written test and oral interview. For the administrators it allows for seamless overview and management of applications and examinations
BVI-SVQA Cross-platform Video Quality Assessment SoftwareELECTRONHTML/CSSJAVASCRIPTPYTHONFFMPEGVMAFEYE-TRACKINGBVI-SVQABVI-SVQA is a software that allows people to prepare and perform experiments for Subjective Video Quality Assessment, do post-processing on the gathered subjecitve data, as well as automatic calculation of objective quality data for each of the videos. Furthermore, visualisation through plot graphs for the subjective and objective data is available independently and combined for comparisson purposes.
Qloud Cloud Quiz ApplicationHTML/CSSJAVASCRIPTPYTHONFLASKMONGODBDOCKERKUBERNETESQloudA cloud based quiz application that lets users connect and compete with friends or an AI in how fast and accurate they can answer different questions. It is built using HTML, CSS and Javascript for the front-end and Python with Flask and MongoDB for the back-end. This application has been containerised with Docker and was deployed on Google Cloud using Kubernetes. The application is currently offline.
Autonomous Car Control System Control System implemented on an FPGAVHDLCELECTRICAL ENGINEERINGAutonomous Car Control SystemA system developed using VHDL and C++ to be programmed on an FPGA, that controls a Multi-Motor Electric Vehicle (EV), as well as its ultrasonic sensors that are used for basic object detection and automatic emergency breaking or slowing down, and an optocoupler inside the tyres that measures revolutions per minute (RPM) and controls the speed of the car. The system can receive and send signals using a Universal Asynchronous Receiver/Transmitter (UART), which can be user controlled, or connected to an AI in order to deliver the autonomous experience.
Disparity Algorithm Optimisation for TI C6678 DSP C-code optimisationCLINEAR ASSEMBLYDisparity Algorithm Optimisation for TI C6678 DSPA optimised version of the Sum of Absolute Differences (SAD) disparity algorithm written in both C and Linear Assembly, meant to be run on the TI C6678 Digital Signal Processor (DSP). The initial version of this algorithm written in C was running at 0.58s. We managed to reduce the runtime and achieve a speedup factor of 20, which resulted in runtime of 0.027s. For comparisson purposes the same code was written in Linear Assembly and resulted in a runtime of 0.038s.
- OpenMP Optimisation of the Jacobi Algorithm C-code optimisationCOPENMPOpenMP Optimisation of the Jacobi Algorithm
An optimisation project for the Jacobi Iterative Method written in C. This was a 2 part project in which the first part was serially optimising the Jacobi Algorithm to run as fast as possible on one core and the second part of this project was taking the optimised code and using OpenMP to enable paralelisation and to apply further optimisations, in order for full utilisation of the CPU. The paralelised code using 16 CPU threads for matrix sizes 4000x4000 and 2000x2000 yielded additional speed up of around 6x and 8x respectively.