Mohammad (Hamudi) Naanaa

San Francisco Bay Area Contact Info
997 followers 500+ connections

Join to view profile

About

Hi there, my name is Hamudi! I'm a Lebanese-born, Ukraine-raised, and Germany-educated AI…

Activity

Join now to see all activity

Experience & Education

  • PortalAI

View Mohammad (Hamudi)’s full experience

See their title, tenure and more.

or

By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.

Volunteer Experience

  • AGI House Graphic

    Moderator

    AGI House

    - Present 1 year 1 month

    Science and Technology

    Moderating the hackathon series at AGI House.

    AGI House is a community of AI builders united by the vision of accelerating humanity's transition to AGI & honoring the greatest Al founders and researchers of our time.

  • Google Graphic

    Core Team

    Google

    - 1 year 1 month

    Education

    Google Developer Student Clubs are university-based community groups supported by Google for students interested in Google developer technologies. By joining a GDSC, students grow their knowledge in a peer-to-peer learning environment and build solutions for local businesses and their communities.

    As Core Team Member, I organize tech talks, educational workshops, and hackathons for club members to grow and diversify our developers community. Excited to learn more? Join us in one of our…

    Google Developer Student Clubs are university-based community groups supported by Google for students interested in Google developer technologies. By joining a GDSC, students grow their knowledge in a peer-to-peer learning environment and build solutions for local businesses and their communities.

    As Core Team Member, I organize tech talks, educational workshops, and hackathons for club members to grow and diversify our developers community. Excited to learn more? Join us in one of our upcoming events and check out the past events at dscmunich.de

    If you have an exciting topic worth sharing and are interested in doing an event with our club as invited speaker, please reach out to me here via LinkedIn or contact us via email found at dscmunich.de!

  • Techstars Startup Weekend Graphic

    Startup Mentor

    Techstars Startup Weekend

    - 9 months

    Education

    Served as technical mentor for startups, focusing on AI research and engineering, product development, and team building and management.

Publications

  • 3D Scene Diffusion Guidance using Scene Graphs

    Visual Computing & Artificial Intelligence Group @ TUM | arXiv

    Guided synthesis of high-quality 3D scenes is a challenging task. Diffusion models have shown promise in generating diverse data, including 3D scenes. However, current methods rely directly on text embeddings for controlling the generation, limiting the incorporation of complex spatial relationships between objects. We propose a novel approach for 3D scene diffusion guidance using scene graphs. To leverage the relative spatial information the scene graphs provide, we make use of relational…

    Guided synthesis of high-quality 3D scenes is a challenging task. Diffusion models have shown promise in generating diverse data, including 3D scenes. However, current methods rely directly on text embeddings for controlling the generation, limiting the incorporation of complex spatial relationships between objects. We propose a novel approach for 3D scene diffusion guidance using scene graphs. To leverage the relative spatial information the scene graphs provide, we make use of relational graph convolutional blocks within our denoising network. We show that our approach significantly improves the alignment between scene description and generated scene.

    See publication

  • Accident Prevention Frontend Framework to Support Autonomous Driving

    Robotics, AI, Embedded Systems Group @ TUM

    · Research project within Providentia++ programme aimed to improve traffic flow and road safety in complex traffic scenarios.
    · Proposed a supervisor infrastructure that builds a real-time virtual twin of the road and detects risk events (jams and accidents) using multiple distributed sensors (LiDaR/optical).
    · Developed a scalable information propagation networking mechanism to distribute warnings to autonomous vehicles.
    · Implemented a demo iOS app to visualize the virtual twin and…

    · Research project within Providentia++ programme aimed to improve traffic flow and road safety in complex traffic scenarios.
    · Proposed a supervisor infrastructure that builds a real-time virtual twin of the road and detects risk events (jams and accidents) using multiple distributed sensors (LiDaR/optical).
    · Developed a scalable information propagation networking mechanism to distribute warnings to autonomous vehicles.
    · Implemented a demo iOS app to visualize the virtual twin and the detected risk events.

  • "Learning with Kernels" - Linear ML models enhancement

    Data Mining Group, Chair of Scientific Computing @ TUM

    · Research paper investigating application of kernels to improve linear machine learning models.
    · Developed a generalized kernelized ML algorithms application pattern and compared the performance with non-linear ML models.

  • Computer Vision Algorithms Acceleration with Hardware Parallelization using SIMD/AVX-512

    Computer Architecture Lab @ TUM

    · Investigated computer vision algorithms hardware acceleration in limited-performance-systems using CPU only with SIMD/AVX-512 registers.
    · Released a paper with implementation details and comparison to a non-optimized reference solution.
    · Our solution achieved a runtime performance improvement of up to 400% (tested on image gamma correction algorithm using AVX-512 vs. pure C solution benchmark).

Projects

  • AGI House Hackathons

    • Series of hackathons in Silicon Valley dedicated to advancing the field of artificial intelligence and building AI products.
    • [07/23] Autonomous Agents Hackathon: Proposed the idea, built a team of 4, developed and presented SyncMate - autonomous social co-pilot AI agent.
    • [08/23] Imagine Hackathon: Moderated the final demos.
    • [09/23 - present] Hackathons moderator.

  • Google DSC Events

    -

    • Organized series of tech talks & educational workshops at Google for a community of 1000+ members.
    • Hosted a high-stakes discussion round with Google's VP (CISO) Phil Venables for the Munich Security Conference.
    • Educated 50-200 participants/session on state-of-the-art AI and ML research topics.

  • piARno - Augmented Reality Piano teaching app

    -

    piARno is an AR piano teaching and practicing app for headsets like Oculus Quest 2. It overlays a physical piano with AR 3D tiles that guide the user's key presses, representing the chosen song.

    In a team of three, I designed the app's software architecture and co-developed in C++ a basic AR engine and the visualization of "falling piano tiles" in AR.

    I also presented the piARno app in the Augmented Reality lecture series at Technical University of Munich (TUM), followed by a…

    piARno is an AR piano teaching and practicing app for headsets like Oculus Quest 2. It overlays a physical piano with AR 3D tiles that guide the user's key presses, representing the chosen song.

    In a team of three, I designed the app's software architecture and co-developed in C++ a basic AR engine and the visualization of "falling piano tiles" in AR.

    I also presented the piARno app in the Augmented Reality lecture series at Technical University of Munich (TUM), followed by a public live demonstration hosted by the AR Research Group at TUM with over 50 test users. As a result, the app received two awards from public voting:
    - Best Technology
    - Best Visuals

    Additionally, I released the app as an open-source project under MIT license for educational purposes to support the emerging AR app developers community.

    Other creators
    See project

  • Ferienakademie 2022 - Decentralized Decision Making in Smart City Infrastructure

    -

    • Research summer school for selected students aimed to explore innovative ideas.
    • Designed the subsystem architecture and implemented autonomous localization based on April Tags using openCV and V2V/V2I communication.
    • Developed a scalable ”watchtower” code deployment pipeline tested on 12 agents as a part of CI/CD using Docker.

    See project

  • JASS Cairo 2022

    -

    • International research project financially supported by McKinsey, ZEISS, and JetBrains. The aim of the project is to investigate various autonomous multi-agent traffic management systems.
    • Led the development of an autonomous driving system tailored for complex traffic scenarios in complex urban environments.
    • Introduced and implemented a geofencing concept to ensure a seamless transition from human control to autonomous navigation as vehicles approach complex roundabouts.

    See project

  • JASS Cyprus 2023

    -

    • International research project (Imperial College London + TUM) aimed to investigate various autonomous traffic management systems.
    • Directed smart traffic control research in a model city and designed the subsystem architecture.
    • Implemented autonomously driving agents from scratch, including sensor data processing pipeline, finite state machine, odometry, computer‐vision‐based lane following with segmentation CNN, and object detection using YOLOv5.

    See project

Honors & Awards

  • piARno — Best Technology Award & Best Visuals Award

    AR Research Group (FAR) @ TUM

    The piARno project (an Augmented Reality piano teaching app for head-mounted AR displays) I co-developed was invited to a public live demonstration hosted by the AR Research Group (FAR) at Technical University of Munich with over 50 test users.

    As a result, the project received two awards from public voting:
    - Best Technology
    - Best Visuals

  • Top 10 in "Analytics Cup" - TUM Machine Learning competition

    Technical University of Munich

    An award for getting into the top 10 (Rank 9 out of 148 4-people teams) in the "Analytics Cup" - a machine learning and data analytics competition - at TUM.

    The goal of this competition was to develop a classification model predicting the success of business offers on a midsize real-world dataset provided by Siemens Advanta Consulting. The provided dataset contained 26.151 anonymized offers made to 8.452 customers in both France and Switzerland and the model performance was measured by…

    An award for getting into the top 10 (Rank 9 out of 148 4-people teams) in the "Analytics Cup" - a machine learning and data analytics competition - at TUM.

    The goal of this competition was to develop a classification model predicting the success of business offers on a midsize real-world dataset provided by Siemens Advanta Consulting. The provided dataset contained 26.151 anonymized offers made to 8.452 customers in both France and Switzerland and the model performance was measured by its balanced accuracy (recall & specificity).

    My responsibilities in the project were:
    * Data preprocessing: data cleaning + dealing with unbalanced data with under-/oversampling + stratified split into train/validation/test sets using cross validation
    * Feature engineering
    * Model building and selection: setup and benchmarking of various machine learning models
    * Hyperparameter tuning of the best performing ML models (Random Forest and Neural Network)

    Additionally, I was responsible for the team building and management - making a roadmap, defining tasks, milestones, and acceptance criteria.

  • Outstanding performance in functional programming and verification

    Technical University of Munich

    An award for "outstanding performance in functional programming and verification" for getting into the top 5% (Rank 27 out of 477 developers) based on 13 weekly challenges in Haskell measuring performance, minimality, provability, and beauty of code.

Languages

  • English

    Native or bilingual proficiency

  • German

    Native or bilingual proficiency

  • Arabic

    Elementary proficiency

  • Russian

    Native or bilingual proficiency

  • Ukrainian

    Native or bilingual proficiency

More activity by Mohammad (Hamudi)

View Mohammad (Hamudi)’s full profile

  • See who you know in common
  • Get introduced
  • Contact Mohammad (Hamudi) directly
Join to view full profile

Other similar profiles

Explore collaborative articles

We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.

Explore More

Add new skills with these courses

See all courses