Fall Term 2020
Winner of the Henri Qais Naccache best fall graduation project prize
Description: CycleMyWaste is a middleman platform solution for the recycling industry in Lebanon. It works by orchestrating and standardizing the way civilians, industries, and the medical field interact with local recycling facilities across the country. It is a unified cross-platform mobile app which works by increasing the exposure of local facilities and accelerating interaction with the user. Users can easily search for recycling facilities based on their location, filter through the facilities, subscribe to an offered service, share a subscription with neighbors, manage and receive pickup-service notifications, and access real-time pickup schedules for every recyclable category.
Team: Maria Daou, Nicolas Kik, Melissa Tanious, Jamil Mahmoud
Advisor: Prof. Wassim El Hajj
Adversarial Machine Learning for Road Signs:
Description: Adversarial attacks pose a serious threat to autonomous vehicles systems. This project presents different adversarial attacks on self-driving cars using adversarial patches on road signs. The project proposed an evaluation of the performance of different algorithms that generate perturbed stop signs. In addition, it showcased the attacks by running them on known classifiers and detectors.
Team: Marina Kayrouz, Miryam Kayali, Hani Alsabe
Advisor: Prof. Mohamed Nassar
Optimized Adversarial Example Detection using ML-LOO:
Description:The project implemented a tuned Multi-Layered - Leave One Out (ML-LOO) approach on the pixels of an image that reduces the computation operations and subsequently improve the time performance, along with a SLIC- K-Means Based Image Segmentation. This method outperformed other adversarial detection methods by detecting perturbations and attacks with significantly improved performance time and a similar detection accuracy.
Team: Bouthaina S. Abdou and Doha AlBared
Prof. Mohamed Nassar
Description: A Slack-like app built from the ground up to enable and facilitate student group studying, designed for education.
Team: Bassel Mawla, Omarou Manaestrli, and Mohamad Khazem Abdul Sater
Advisor: Prof. Said Raheel