Intelligent Systems Lab (ISL) was founded in 2018 within the Department of Electronics and Communications Engineering at the Arab Academy for Science, Technology, and Maritime transport. We conduct research in intelligent and autonomous systems, image and video processing and computer vision, image synthesis and immersive technologies, signal analysis, and data science.Since its establishment, the ISL members have:-Maintained a close connection with the industry through participating in R&D projects and through collaborations with local startups. -Participated and conducted world-class research on a variety of topics including computer vision, internet of things, and bio-signal analysis. -Started the first DL Meetup in the AAST through which they raise awareness, build capacities, and networks DL enthusiasts and tomorrow’s stars in the AAST.

Working in a team of researchers to create practical tools for software synthesis and repair by applying statistical methods on big data. Created a version controlled database from the 1000 most active C/C++ projects on GitHub. Integrated the Apron numerical abstract library in the LLVM/Clang compiler infrastructure to efficiently extract the semantic differences for all projects' history code updates. Utilized machine learning (deep learning) to discover a statistical model of the frequently patched vulnerable code patterns and the required code patch to fix them. Multithreaded programming was used to achieve high scalability and low-latency.

Introduced a quantitative model to estimate the degree of information leakage from power side-channels through static analysis of the cryptographic source code. Statistical analysis of measured power signals in running cryptographic devices matched accurately our proposed quantitative model.Developed tools to formally verify cryptographic software and embedded hardware code against vulnerabilities of power side-channel and fault-based attacks. Furthermore, developed other tools to formally synthesize, from an unsafe code, an equivalent code but secure and optimized. The synthesized code was shown to be more concise (optimized) than the code written by cryptographic experts.Created a tool to optimize C/C++ embedded software in order to overcome the relative limited hardware resources in embedded systems. The tool was integrated in the LLVM/Clang compiler infrastructure and repeatedly invoked an SMT-solver to synthesize highly optimized software through inductive synthesis. Received the FMCAD best paper award for this work which was described it as “an innovative and scalable solution”.

Teaching courses: Digital Signal Processing (DSP), Analog Signal Processing, Measurements and Instrumentation, and Digital VLSI.