Personal Portfolio
Profile Pic

Rajan Shukla :

As a postgraduate from the prestigious IIT Mandi, I have received a strong foundation in ML and i am always looking for opportunities to further develop my skills and knowledge. I am confident in my ability to tackle any challenge that comes my way due to my passion for staying up-to-date with the latest technologies. My skills in Python, Tensorflow, and computer vision make me well-equipped to handle a wide range of projects and tasks. I also have experience in SQL, Excel, and Linux, which adds to my diverse skill set as a developer. I am able to use these skills to effectively analyze and manipulate data, as well as work with operating systems and other tools. In addition to my technical skills, I also have knowledge of Docker and Kubernetes, which allows me to easily deploy and manage applications in a production environment. This enables me to deliver reliable and scalable solutions to my clients and employers. To get a better understanding of my capabilities, you can check out my live project, which showcases some of my work in action. I am always looking for opportunities to challenge myself and learn new things, and I believe that my skills and experience make me a valuable asset to any team.

Skills

Over the course of my M. Tech I have focused on Acquiring the skill on ML/DL till the deployment. I have listed the sills, Languages & Tools that i have used for creating the project listed in next section

My Work

content

SDA-WEBTOOL(LATESTPROJECT)

I developed a web-based, on-demand data analytics tool for analyzing the NWH dataset for features such as temperature, surface pressure, and wind speed. This tool is an interactive web application that allows users to select the dataset and view spatial plots and traditional plots. In addition, users can select a specific time window to focus on a particular time period.Optimized code and algorithms to efficiently handle large NWH dataset. Applied skills in machine learning, web development, and data analysis to successfully develop and deploy interactive web application. Receiving positive feedback from users.

Live link
content

White Board Camera - Computer Vision

"White Board Camera" was developed using Python and OpenCV. The aim of the project was to correct skewness and uneven brightness from a camera feed through the use of image processing techniques. The hardware platform used was a Raspberry Pi 4, and socket programming was utilized to establish a connection between the camera and the Raspberry Pi. This allowed for real-time capture of the camera feed and application of the image processing techniques. The project was recognized with a 3rd place win at the Engineering Day competition at the authors' campus. Skills in Python, OpenCV, and socket programming were applied in a real-world setting through this project."

Youtube Demo
content

Fall Estimation Diagonostic System.

In this study, we developed a TinyML-based diagnostic system for estimating muscle failure in older adults. Our system utilizes sEMG and IMU sensors to capture data on muscle activity and movement, and applies machine learning algorithms to analyze this data in real-time. By providing early warning signs of muscle failure, our system has the potential to help older adults maintain their independence and prevent falls and other injuries. Our aim is to show that our system is accurate and reliable.

GitHub
content

Surface Crack detection (DIP)

In this work, i present a method for detecting cracks on the surface of walls from images using simple digital image processing techniques. Our method utilizes Otsu thresholding, erosion and dilation, and histogram equalization to identify and highlight cracks in the images. We evaluated the performance of our method using a dataset of images and achieved an accuracy of 82% in Matlab. We also developed a GUI to allow for easy usage of our method. Our results demonstrate the effectiveness of our approach and its potential as a practical tool for detecting cracks on the surface of walls. The code for our method is available on GitHub (link provided).

GitHub
content

Human Pitch Estimation

Pitch is defined as the fundamental frequency of a spoken signal and is a major auditory attribute of musical tones, along with duration, loudness, and timbre. The aim of this project is to use the auto-correlation function to identify the pitch of discrete spoken speech. The audio signal is divided into overlapping frames, and auto-correlation is performed on each frame to derive the pitch estimation. Voiced and unvoiced samples are first separated based on energy content, and then the proposed algorithm is applied. The effectiveness of the proposed method is demonstrated through experimentation.

GitHub

My Blogs

Advance in Python Sorting in incredible ways!

medium

Click

Top 10 Python Advance Techniques You Should Be Using In 2023.

medium

Click

Why Statistical test is important for the linear regression model.

medium

Click