I am a Senior ML Research Engineer at Samsung AI Center in Toronto, where my research focuses on multimodal AI, working at the intersection of vision and language. Previously, I completed my Master’s in Machine Learning at Mila and University of Montreal with Dr. Sarath Chandar, where I mainly worked on lifelong and incremental learning systems. I hold a Bachelor’s in Aerospace Engineering from Zewail University of Science and Technology. Check out my personal CV for more details!
Kalliopi Basioti, Mohamed A. Abdelsalam, Federico Fancellu, Vladimir Pavlovic, Afsaneh Fazly
ECCV 2024
A novel approach for controllable image captioning using structured semantic representations. The method automatically generates focused and diverse image descriptions by leveraging Abstract Meaning Representation (AMR) graphs, improving both caption quality and controllability.
Mohamed A. Abdelsalam, Samrudhdhi B. Rangrej, Isma Hadji, Nikita Dvornik, Konstantinos G. Derpanis, Afsaneh Fazly
ICCV 2023
A generative deep learning model to predict next steps in procedural videos using natural language. The model excels in generating diverse and plausible outcomes through zero-shot transfer from text to video, setting new benchmarks on YouCookII.
Mohamed A. Abdelsalam, Zhan Shi, Federico Fancellu, Kalliopi Basioti, Dhaivat J. Bhatt, Vladimir Pavlovic, Afsaneh Fazly
CoNLL 2022
A novel method for visual scene understanding using Abstract Meaning Representation (AMR) to create linguistically informed visual AMR graphs for higher-level semantic concepts.
Mohamed A. Abdelsalam, Mojtaba Faramarzi, Shagun Sodhani, Sarath Chandar
CVPR 2021
A setup and benchmark for evaluating lifelong learning models in more dynamic and realistic scenarios.
Prasanna Parthasarathi*, Mohamed A. Abdelsalam*, Joelle Pineau, Sarath Chandar
SIGDial 2021
An exploration of semantic loss as an auxiliary objective to improve diversity in dialogue response generation, showing particular effectiveness for smaller datasets.