Skip to content
View Ali-Shariati-Najafabadi's full-sized avatar

Block or report Ali-Shariati-Najafabadi

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this userโ€™s behavior. Learn more about reporting abuse.

Report abuse

Typing SVG

profile views followers


๐Ÿ‘จโ€๐Ÿ’ป About Me

๐ŸŽ“ Computer Science Student @ TU Darmstadt
๐Ÿ’ผ Software Developer @ Fraport AG (Aug 2023 - Present)
๐Ÿ”ฌ Bachelor Thesis: Multimodal Deepfake Detection @ Fraunhofer SIT ๐Ÿ“ Based in Darmstadt, Germany

const ali = {
    code: ["Python", "JavaScript", "Java", "TypeScript", "C++"],
    technologies: {
        frontEnd: ["React", "Vue.js", "HTML", "CSS", "Tailwind"],
        backEnd: ["Node.js", "Flask", "FastAPI"],
        databases: ["PostgreSQL", "MongoDB", "MySQL"],
        devOps: ["Docker", "Git", "Azure DevOps", "CI/CD"],
        testing: ["TestComplete", "ReadyAPI", "Gherkin", "Selenium"],
        ai_ml: ["PyTorch", "TensorFlow", "Hugging Face", "OpenCV", "Scikit-learn"]
    },
    currentFocus: "Deep Learning & Test Automation",
    funFact: "I debug production code and deepfakes ๐Ÿค–"
};

๐Ÿ› ๏ธ Tech Stack


๐Ÿš€ Featured Projects

System Security Lab, TU Darmstadt

Development of a performant real-time pipeline for voice and face conversion with client-server architecture and user-friendly GUI.

Tech: Python โ€ข PyTorch โ€ข Computer Vision โ€ข GFPGAN โ€ข InsightFace โ€ข Diff-HierVC

Highlights:

  • Real-time audio and video processing
  • Client-server architecture
  • GUI for easy interaction

๐Ÿ”’ Tool Support for Sustainable Protocol Security

SEEMOO, TU Darmstadt

Full-stack web application for analyzing LoRaWAN security vulnerabilities from 100+ research papers with interactive visualization.

Tech: Python โ€ข React โ€ข Security Analysis โ€ข CLI Tools โ€ข LaTeX

Highlights:

  • Analysis of 100+ security papers
  • Interactive graph visualizations
  • Automated LaTeX table generation

System Security Lab, TU Darmstadt

Training of a BLIP-2 model on GoodNews dataset (15k images) for context-aware image captioning.

Tech: Python โ€ข PyTorch โ€ข NLP โ€ข Transformers โ€ข BLIP-2

Results:

  • 23% improvement over baseline (BLEU score)
  • Trained on 15k images
  • Context-aware caption generation

๐ŸŽ“ Multimodal Deepfake Detection

Bachelor Thesis @ Fraunhofer SIT

Development of a deep learning system for detecting and temporally localizing audio-video manipulations in deepfakes.

Tech: PyTorch โ€ข CNNs โ€ข 3D-CNNs โ€ข Wav2Vec โ€ข HuBERT โ€ข Transformers

Highlights:

  • Audio-video inconsistency detection
  • Temporal localization of manipulations
  • Evaluation on ACM Grand Challenge dataset

๐Ÿ“Š GitHub Statistics

streak stats

๐Ÿ† GitHub Trophies


๐ŸŒ Connect With Me


๐Ÿ’ผ Work Experience

current:
  role: "Working Student Software Development & Test Automation"
  company: "Fraport AG"
  location: "Frankfurt am Main, Germany"
  period: "Aug 2023 - Present"
  highlights:
    - "Implementation of 60+ end-to-end tests with TestComplete and JavaScript (Gherkin)"
    - "Reduced manual testing time by 60%"
    - "Development of 70+ automated API tests with ReadyAPI"
    - "Integration into CI/CD pipeline for Kafka message validation"
    - "Error documentation and analysis in JIRA, task management in Azure DevOps"
    - "CI/CD pipeline optimization in agile team, regular participation in sprint reviews"

๐ŸŽ“ Education

B.Sc. Computer Science - Technical University of Darmstadt
๐Ÿ“… Oct 2021 - Present

Bachelor Thesis: Detection of Multimodal Deepfakes @ Fraunhofer SIT

  • Development of deep learning system for audio-video manipulation detection
  • Combination of CNNs/3D-CNNs and Transformer architectures (Wav2Vec, HuBERT)
  • Evaluation on ACM Grand Challenge dataset

๐Ÿ’ก Current Focus

  • ๐Ÿ”ญ Working on multimodal deepfake detection for my bachelor thesis
  • ๐ŸŒฑ Learning advanced deep learning architectures and transformer models
  • ๐Ÿ‘ฏ Open to collaborate on ML/AI projects and open-source contributions
  • ๐Ÿ’ฌ Ask me about Python, React, PyTorch, Test Automation, Computer Vision, NLP
  • โšก Fun fact: I can debug both production code and AI-generated deepfakes!

๐Ÿ“ˆ Key Achievements

  • ๐Ÿ† 60% reduction in manual testing time through test automation
  • ๐ŸŽฏ 23% improvement in BLEU score for context-aware image captioning
  • ๐Ÿ“Š 130+ automated tests (60+ E2E, 70+ API tests) in production
  • ๐Ÿ”ฌ Research on multimodal deepfake detection at Fraunhofer SIT
  • ๐Ÿ› ๏ธ Full-stack development of security analysis tools for 100+ research papers

๐ŸŒ Languages

  • ๐Ÿ‡ฉ๐Ÿ‡ช German: Advanced (Professional working proficiency)
  • ๐Ÿ‡ฌ๐Ÿ‡ง English: Advanced (Professional working proficiency)
  • ๐Ÿ‡ฎ๐Ÿ‡ท Persian: Native speaker

snake animation

๐Ÿ’ซ Show some love by starring repositories you find interesting!

Popular repositories Loading

  1. Real-Time-Deepfake-Pipeline Real-Time-Deepfake-Pipeline Public

    Real-Time Deepfake Pipeline

    Python 24 8

  2. Generate-Context-Aware-Captions-from-Photos Generate-Context-Aware-Captions-from-Photos Public

    Context-Aware Image Captioning with BLIP-2

    Python 3

  3. project project Public

    Java

  4. Ali-Shariati-Najafabadi.github.io Ali-Shariati-Najafabadi.github.io Public

    Mein persรถnliches Portfolio

    JavaScript

  5. Ali-Shariati-Najafabadi Ali-Shariati-Najafabadi Public