Graduation projects of participants of the senior bootcamp “codeIIm” (July 2024)
Our team organizes bootcamps on artificial intelligence for teenagers. We teach kids to create their own projects using neural networks, teach programming, mathematics and data analysis in the process of working with AI. We also build a community of teenagers (and sometimes even adults). All updated information about projects and more is on our Telegram channel “codeIIm sandbox”!
During the year, our team organizes several bootcamps designed for different levels of knowledge. Our bootcamps are off-site shifts where students live and study together. During the week, participants master fundamental topics in artificial intelligence at seminars and lectures, and at the end, they implement their own project under the guidance of teachers and mentors, applying the knowledge they have gained in practice. In addition to studying, in the evening, participants play sports and board games, which helps them relax and prepare for the next day.
The last shift ended at the beginning of July, and it is about this shift that we want to tell you.
The Senior Bootcamp took place from June 30 to July 10 and was designed for schoolchildren in grades 9-11 who already had deep knowledge of programming and mathematics. This shift was divided into three tracks: CV, NLP and ML, within which the participants made their final project.
What is the final project? It could be a solution to a real case from top companies or a personal project, the topic of which the students themselves suggested. This year, Tinkoff, Wildberries and VK shared their cases.
The projects turned out to be complete and well-packaged, so we really want to tell you about each of them!
CV Track
Team “3\4”
Participants: Aldag Alsou, Budennaya Maria, Bolshchikova Yaroslava, Gainullina Nastya
The girls solved a case from Wildberries, the goal of which was creation of an anti-fraud system for further implementation in an online store. This system will help to identify fraudulent reviews and prevent disclosure of personal data, thereby strengthening customer trust.
GitHub: https://github.com/bbalsu/antifraud
Team “APS”
Participants: Chekueva Alima, Latypova Yulia
The APS team has developed a neural network that helps the Wildberries marketplace classify defective and non-defective backgrounds of product cards. This business task will help sellers sell goods on the platform more efficiently and quickly.
GitHub: https://github.com/alickqs/Codaim_BGdetector
Team “Center for the Study of Asian Norms (CIAN)”
Participants: Gunaev Rostislav, Duvanov Luka, Susak Anna
This team proposed to make life easier for “cleaners” and translators who process the original text and translate it into another language, in our case from Japanese to Russian. The neural network itself fills the “bubble” with the text and automatically inserts the new translated text.
GitHub: https://github.com/AnnaSusak/textInBubbleDetecting
Team “OpenAiAi”
Participants: Krupnov Pavel, Erguchev Alexander, Ionchikov Alexander
The guys from the OpenAIAI team developed a neural network that will help analyze the audience at conferences. Why? Many events are ineffective and boring, and the organizers cannot understand the reason for the loss of the audience. This neural network will help recognize people's mood and adjust further events depending on the analysis results.
GitHub: https://github.com/krup4/face-feature-recognition
NLP/ML Track
Team “Decepticons”
Participants: Mikhail Tsypchenko, Ilya Krom, Andrey Boriskin, Nikita Nefedov, Miroslav Adamenko
The Decepticons team took a case from T-bank and developed a neural network that determines a person's lisp by recording their voice. A good case that will increase the effectiveness of the security system and help with biometric data recognition.
GitHub: https://github.com/Vex1cK/Decepticon-s-project
Team “42”
Participants: Sukhanovsky Ignat, Spirin Konstantin, Konukhov Artem, Fesenko Mikhail, Malakhov Yaroslav
How do you like it? trading botwhich will predict the movement of MOEX quotes? Team “42” took up the development of this model. In addition to processing numerical data, the guys built in text analysis in the form of news to improve the accuracy of predictions.
GitHub: https://github.com/iwance/sber_stocks_prediction
Team “Hedgehog neurons”
Participants: Belyaev Alexander, Kharitonov Georgy, Khvatov Sergey, Veselov Danila, Irkhin Andrey, Golyshev Yuri
If you have a project that could potentially win some funding, this bot will definitely help. The guys from the Hedgehog Neurons team created modelwhich will predict the outcome of a grant application based on the text. A very useful thing, especially at the initial stages!
GitHub: https://github.com/Sashakrem8320/Telegram-Bot-Grant-forecasting
Team “sigmaminded”
Participants: Ivan Gronsky, Ignatiy Arkhipov, Grigory Gavrilov, Miron Labus
For people who create content, the project of the team “sigmaminded” is especially useful. The bot analyzes the relevance of the video and predicts the number of views, and also gives recommendations for improvement. This will significantly save the efforts and resources of all creators.
GitHub: https://github.com/Sgsaram/ai-podcast
Team “egorcrid”
Participants: Ivan Sazanov, Dmitry Yushkov, Lev Klyuev, Arina Sadchikova
Another group of people who like to predict the situation on the stock market. The case from Sber Investments at the AI Challenge turned out to be very popular, and most importantly, successfully completed. Here, team members predict quotes using news analysis.
GitHub: https://github.com/iwance/sber_stocks_prediction
We are very proud that the guys were able to create well-thought-out and packaged projects in such a short time. Everyone has potential for further development and expansion. The “kodIIm” team believes in you!
We are waiting for the next shifts!
