I am Samrat Sahoo, a BS Computer Science at Georgia Institute of Technology. My primary interests are within the fields of artificial intelligence, application development, and enterpreneurship. More specifically though, I have a real interest for machine learning, computer vision, mobile app development, and company building. As I continue my journey within the world, I believe I will be able to achieve a lot due to my passion for computer science and integrating within the world to make a change.
- Software Engineering Intern - Working on Computer Vision Applications
Grades 11th & 12th
- Research Assistant - University of Texas at Arlington
Grades 11th & 12th
- Research Mentee
- Facial Recognition Based Attendance Tracker
Grades 11th & 12th
- Using Artificial Intelligence to Accelerate Accessible Education
Georgia Institute of Technology
Rick Reedy High School
University of Dallas
Samrat Sahoo doesn't want to be a politician. Despite that, the Reedy High School rising senior wanted to participate in a program that postures itself after government operations.
Texas Boys State, an American Legion participatory program, allows high school students to become a part of their local, county and state government operations, according to the program website.
2020 - 2021
First Robotics Competition
Rick Reedy High School
Academic All Star Award
National Merit Commended Scholar & National AP Scholar Award
2019 - 2020
First Robotics Competition
District Championship, Rookie All Star, Highest Rookie Seed (x2), Rookie Inspiration Award
Reedy High School
Computer Science School Year Award, Academic Excellence Award
2018 - 2019
Boys Scouts of America
Eagle Scout, Eagle Palms Gold and Silver
3rd Place Team Award (District)
Physics Virtual Challenge
6th Place 10th Grade Physics (District)
During my research internship under Professor Won Hwa Kim, I researched a variety of subjects including but not limited to: generative adversarial networks, conditional generative adversarial networks, convolutional neural networks, and machine learning algorithm evaluation methods. As this internship is ongoing, as I continue my work with professor Kim, I would like to utilize neural networks and generative adversarial networks in order to help either denoise medical images or create medical images froms scratch.
During my research menteeship with software developer Trey Blankenship, I learned about the foundations of computer vision and machine learning. I then applied the knowledge I learned to create a facial recogniton based attendance tracker which has several features including a connection to google sheets and microsoft excel, liveness detection, online/offline mode, dynamic face addition, and so much more. I used the flask framework to build the beautiful user interface and I used the Keras deep learning library to create a convolutional neural network for the liveness detection. The core program was written in Python.
I took part in an MIT Lincoln Lab program known as Beaver Works Summer Institute (BWSI). This was during the COVID-19 pandemic when the contact tracing was vital to minimizing the cases from COVID-19. However, because manual contact tracing was not very effective, we implemented a Bluetooth based contact tracer with Raspberry Pis. My specific project was evaluating the recieved signal strength indicator based on thin obstructions. To design the project, I collected CSV files for a scanning raspberry pi that would pick up the advertiser from 2, 4, 6, 8, and 10 meters. I would repeat this except with a thin box acting as the obstruction. The CSV files included the RSSI strength which would then be plotted. I would then use proximity detection algorithms to compare the predicted distances based on RSSI vales for both the box and non-box experiments.
This facial recognition based attendance tracker is a piece of open source software that has gotten over 100 downloads as per GitHub statistics. This software has several features including Google Sheets API integration, Microsoft Excel integration, Liveness detection, dynamic face addition, online/offline mode, and other auxillary features. The Google Sheets API/Microsoft Excel integration allows for the attendance to be displayed in a visually appealing way. The liveness detection prevents spoofing and was written with the Keras deep leanring framework. The dynamic face addition allowed users to add faces on the fly. The online offline mode checked the status of the internet and toggled google sheets or microsoft excel depending on the if there was internet or not.
This project was creating during the coronavirus pandemic where many children are forced to stay at home and receive lectures and textbook chapters from their teachers with a lack of real projects to instill content. As a high schooler myself, I would really like a way for us to take pages from our textbooks and transform them into a concise summary with questions generated in order for me to truly grasp the concepts I am learning. I created a web and mobile application that allows the user to input a piece text, image, or audio file from their device and our program will process the file and extract a summary, generate supplemental questions as a quiz, and create flashcards over key terms.
Videos Created by Artificial Intelligence • 10 Videos Daily • Brush up or Learn in-depth about ANY topic with programmatically chosen graphics. A channel that leverages natural language processing and Wikipedia to automatically create educational videos.
The Classified Mobile application was created in an effort replace previous iterations of a Home Access Center Mobile Client. In order to prevent a corporate monopoly, I co-created Classified. It was created using a Flutter and Dart frontend with a Python Restful API backend. The purpose of this application was to allow students to more easily check their grades that are from home access center. This would allow students to navigate a beautiful user interface while also being able to use HAC at their fingertips without any trouble. The Classified Application was discontinued after FISD announced a switch to a new grading platform; however, the API still exists and can be reused for other districts with a few small changes.
During my time as a leader within the troop 4018, I held 3 primary positions. I was first the scribe for 12 months where I would handle troop organization and meetings. During this time, the troop was extremely organized and functioned flawlessly. I was then the webmaster and would consistently update the troop website. I was finally the Assistant Senior Patrol Leader and and took care of the duties of the senior patrol leader when he was absent. Through the culmination of these experiences, I was able to direct a completely scout drive project while also obtaining the rank of Eagle Scout and several Eagle Palms. Even after my leadership ended, I contined to stay active wihtin the troop by informing others of my eagle scout journey and helping younger scouts achieve the same feats I did.
During my time at Boys State I was elected to the Rules Platform Committee as my city's representative and I was also appointed to the governor's committee on the last day as the Military Preparedness Comissioner. During my time at Texas Boys State, I learned a great deal about the political system. At Texas Boys State we would run for positions for our respective political party (nationalists or federalists). The Rules Platform Committee decided the rules for the election. Because of my work in the Rules Platform Committee, the Nationalist party (my party) took 7 out of 8 state office positions resulting in a large win for us. 2020 was the first time in the 85 year history of the program that it was held virtual. Because of this, we had to prepare speeches that would match the medium of communication. Because of my hard work and efforts at Boys State, Star Local Media featured me and I was the main story on the front page of the Frisco Enterprise.
Building robots always seemed like something I would not be able to do until very late in my life. That is why I joined the Radicubs Robotics(A FIRST Robotics Competition Team) as one of the founding members. Throughout my 3 years within the Radicubs, I would have the chance to explore my passions within software development and enterpreneurship. The Radicubs has served as an outlet for students in Frisco to explore their passions. Throughout my journey with the Radicubs, I started out as a software developer. My hard work and dedication to the team was recognized and I was promoted as the Junior for the following year in a unanimous vote. I continued my dedication and eventually became the primary senior captain of the Radicubs. Over the 2 years we were as a team we have formed over 10 corporate relationships and have raised $40,000. We have also built our legacy through partnering with the community around us and winning several awards including the Entrepreneuship award, Highest Rookie Seed (x2), Rookie Inspiration, and the Rookie All-Star Award. I plan to continue our streak this upcoming year as I lead the team as our senior captain. Due to my love for progress over perfection, we have so far 2 months ahead of where we were last year. If we continue this streak, we have good chances for success for not only this year but for future generations to come.
As someone who has always had a passion for math, Mu Alpha Theta was the perfect club for me. This was the first club I joined my freshman year. During my freshman year, the club had a mere 7 members and I was one of the most active. As my dedication shined, I took on the role as the Vice President of the club. As Vice President I took on most of the duties of the other officers. This allowed us to have more engaging and meaningful meetings. By the end of my Junior Year, the club had around 25 members. Currently, I am seriving as the President of the club and I plan to increase membership, make meetings worth more, and make the club a more enjoyable place in general.
I am the Founder of Project Codelaborate, a programmming education organization located in Frisco, TX. Project Codelaborate has partnered with a local middle school to spread the love for programmming. We have also partnered with 6 organizations from texas in order to help make our platform better and reach more people. We also have recruited volunteers nationally in order to expand our online curriculum which consists of over 30 programmming videos for Java and Python. Our YouTube channel as amassed over 300 views within its time.
For my Eagle Project, I did a service to the Chambersville Tree Farm. After about 4 months of planning, preparation, and doing, my eagle project would be completed. This project was known was Trails Remade. Trails Remade was a project where I would help reconstruct a 240 square feet trail after it had been infected by termites. In order to do this, I had a multi-step process. The first step was to get wooden boards that were actually donated by the tree farm. I would then need use a nice wooden color to stain the boards. I then needed to spray some termite killer to remove the termites from the area. After that was completed, I needed to put the boards around the actual train in a 30 feet by 8 feet configuration. The final step would be to elevate the trail so the change in elevation from boards to the ground was 0. I did this by using some extra dirt they had at the bottom of a hill. This project was completely led and planned by me and I had gotten more than 20 scouts and adults involved for the entire process.
In the summer of 2020, I took part in Harvard's famous CS50 class known as Introduction to Artificial Intelligence with Python. I learned a lot of new concepts during my time when taking that class including but not limited to: graph search, propositional logic, Bayesian Networks, constraint satisfaction, reinforcement learning, recurrent networks, and n-gram models. These concepts helped build my fundamentals in many areas of artificial intelligence. Previously, most of my knowledge was in the field of machine learning, natural langauge processing, and computer vision but now I was famaliar with all the subsets in AI. Additionally, my Python programming skills grew as a result of taking this class, hence, contributing to my overall success in other ventures.
During the summer of 2020, I took part in Cornell Univeristy's Robotics and AI program in order to gain a greater appreciation, interest, and knowledge for robotics. During my time there, I created a robot that synced up with other robots in the vicinity and would transmit a bluetooth signal to those robot which would cause all of the robots to dance in sync. I also learned a lot more about the electrical aspect of a robot because I face hardships regarding my electrical board. Through these hardships, I perservered and was able to push onto further tasks by finding workarounds. I also learned a lot about the programming aspect as I learned a new programming language tha I could apply to other projects. In these 2 weeks, I went from a robotics amateur to someone who could construct a robot on his own.
In the summer of 2020 - 2021, I took part in a virtual summer program at Rose-Hulman Institute of Technology to further explore my interests in the fields of Physics and electrical engineering. During the program, we did several experiments with pendulums, circuits, and bridges. With pendulums we calculated the acceleration of gravity. With circuits, we built an electric motor that was powered by an AA battery. With bridges, we measured their strength through looking at the amount of weight it could sustain. Throughout this whole program we looked at a very important science concept known as uncertainty but more importantly the different types of uncertainty such as random uncertaintiy, resolution, absolute uncertainty, and more. In taking part in Rose Hulman's creation crates, I was able to clarify my interests within the field.