Page 184 - tech fest 2025 ב״ש חוברת תקצירים

184
Deepfake in Videos
SE-D-11
Omer Assenheiner; omeras2@ac.sce.ac.il Oscar Aizenshtat; oscarai@ac.sce.ac.il Yuval Soto Korenstein; yuvalso@ac.sce.ac.il
Advisor: Ms. Alona Kutsyy
SCE - Shamoon College of Engineering, Be'er-Sheva
The project focuses on detecting deepfake videos by developing a new model that combines ResNeXt and LSTM architectures. The goal is to create a robust and efficient system capable of identifying deepfakes in real-world, “in-the-wild” conditions. The model is trained on a diverse and challenging dataset to handle various video qualities, lighting conditions, and facial expressions. By integrating spatial and temporal features, it captures both image-level and sequence-level patterns typical of deepfakes. This approach enhances accuracy and generalizability. The project addresses the increasing need for reliable deepfake detection tools, especially as manipulated media become more advanced and harder to detect.
Keywords: dataset, deepfake detection, lstm, resnext
Gaia Gaia: An Intelligent Trip Planner
SE-D-12
Dor Yubiler; doryubiler@gmail.com Shani Riven; shaniriven@gmail.com
Advisor: Ms. Alona Kutsyy
SCE - Shamoon College of Engineering, Be'er-Sheva
Gaia is an intelligent trip planner that creates personalized itineraries based on user input. Travelers provide their travel dates, destinations, and interests, such as museums, art, and cafés, and Gaia delivers a tailored list of top attractions and local insights. Using advanced AI, Gaia crafts precise search queries to retrieve detailed information from trusted sources, ensuring that recommendations are relevant and current. The application also checks for existing trip data, allowing users to update or keep their previous plans. By combining real-time data with personalized recommendations, Gaia simplifies the planning process and enhances the travel experience for every user. Gaia empowers travelers to explore new destinations confidently, ensuring every journey is seamlessly planned and tailored to individual tastes.
Keywords: ai, itinerary, personalized, travel