Hi, I'm Brian Marcellino.

M. Aldiki Febriantono, Nasywa Hilmi Cahyono, Brian Marcellino

Compared lightweight architectures (MobileNetV1, MobileNetV2, MobileNetV3, MnasNet, TinyNAS) on a custom Poncokusumo apple-orchard dataset (6 classes, ~10,668 images after augmentation). Reports accuracy, precision, recall, F1, MIoU, and training time, then validates the best model in a mobile application.

M. Aldiki Febriantono, Brian Marcellino, Nasywa Hilmi Cahyono

Developed a lightweight, mobile-deployable deep learning model for apple disease classification, optimized for on-device inference on resource-constrained devices.

Brian Marcellino, Nasywa Hilmi Cahyono, M. Aldiki Febriantono

Comparative analysis of deep learning architectures (InceptionV3, ResNet50, EfficientNetB0, MobileNetV2) for apple leaf disease classification across multiple public datasets, achieving a peak testing accuracy of 98.89%.

Brian Marcellino, Nasywa Hilmi Cahyono, M. Aldiki Febriantono

Developed a deep learning framework using InceptionV3 with the Adam optimizer to classify apple fruit diseases. Achieves 97.15% test accuracy across five categories including Black Pox and Black Rot.

Brian Marcellino, Nasywa Hilmi Cahyono, Prash Trisula Ajinata, Anita Rahayu

Researched and compared two forex trading robots (BotAGI and Dark Rea) on the EUR/USD pair, revealing profitability differences and contrasting risk profiles.

Alvin Ariyanto Putra, Brian Marcellino, Bryan Fernando Kurniawan Suhartono, Wildan Rizkia Noor Akhsal Hermawan, Frihandhika Permana

Developed an AR-based application enabling real-time 3D property visualization and construction tracking, using marker-based detection to anchor digital models in physical space.

Muhammad Hafizh Raihan Daniswara, Frihandhika Permana, Ahmad Hafidz, Brian Marcellino

Conducted a detailed evaluation of Duolingo's user experience using the User Experience Questionnaire (UEQ), highlighting strengths in attractiveness and efficiency while identifying improvement opportunities in novelty.