In this project, supported by Tubitak and completed as a graduation project, I extract the person and then the clothing information from CCTV recordings using advanced instance segmentation and image classification techniques, and I save this information for future security use.
Further Reading...
In this project, which received great interest at Teknofest, I extract advanced features of the selected artist's source image to teach the style of the image to my model, and then I apply this style to the user image obtained through the interface with a background added. The person on the screen transforms into a work of art within a few seconds.
Further Reading...
In this project, where I did not use any additional libraries other than Numpy, I embark on a journey to the extremes of image processing and apply a unique filter to the images I provide as input.
Further Reading...
Can there be convolution in PLC? Don't say that. In this project, I simulate an Amplitude Modulator (AM) in the Codesys environment and demodulate the high-frequency signal I produced as if it had reached the receiver to revert it to its original state. I demonstrate all of this through the HMI.
Further Reading...
In the solution development process initiated after the Kahramanmaraş earthquake, this project, which I lead, extracts address information from aid posts on social media using advanced NLP techniques and then marks it on a map to provide information to aid organizations.
Further Reading...
Words need to have a numerical representation for computers to understand them. The concept of Word2Vec comes into play here. In this approach, where we represent words with vectors, I tried to find an answer to the question of whether I can classify the vectors of sentences using clustering logic on an unsupervised dataset.
Further Reading...
In this project, where I developed a movie recommendation system with different approaches, I aimed to present content that users would love and to provide new suggestions.
Further Reading...
In this project, where I provide descriptions for images through Vision Transformer and LLM models, I input the content of social media content creators into the system via web scraping and then perform class analysis of their content.
Further Reading...
In this project, I am deploying our developed clothing classification and querying application as a web application to the internet using Docker and GCP.
Further Reading...
In this project, I am transferring my ESP32 board to the IoT application environment of Blynk and then operating the touch and IR sensors and the stepper motor.
Further Reading...
In order to train a model using a dataset of images representing letter and digit characters using CNN architectures, I used grid search to do hyperparameter optimization for this project. I also analyzed the accuracy data and examined at the input and output feature maps.
Further Reading...
In this project, I created a site that features user authentication, allowing members to create their own movie lists, start threads and share posts in the forum, interact with other users, and vote on films.
Further Reading...
I analyzed the location data and quantity information of the Emergency Gathering Areas in İzmir, sometimes in an interactive manner, using various visualization methods.
Further Reading...
In this study, I made comparisons regarding the amount of grapes produced by countries and the quantity of wine they produce, visualizing this data on tables and maps.
Further Reading...