About Me
I architect, train, and deploy multimodal AI systems spanning 2D/3D Generative AI, LLMs, and computer vision. As a Research Software Engineer and Data Scientist with a Ph.D. in Computer Science, I bring 3+ years of experience delivering AI/ML solutions in heavy industrial operations, combining strong foundations in statistics, machine learning, and predictive analytics with proven business impact.
My expertise covers end-to-end ML pipelines: from data extraction, feature engineering, and statistical modelling to model development, deployment, and monitoring. I have delivered measurable improvements in operational efficiency (130% throughput gains, 20% error reduction) across industrial robotics, telecommunications, and AR/VR applications. Proficient in Python, SQL, PyTorch, and TensorFlow, with hands-on experience in Databricks, Apache Spark, and AWS for scalable data processing and ML operations.
I specialize in diffusion models (DDPMs, Flow Matching) for image and video generation, with hands-on experience in multi-billion-parameter DiT architectures (Wan-Video, CogVideoX) and UNet-based systems (Stable Diffusion, Stable Video Diffusion). My technical expertise includes multi-GPU distributed training, custom CUDA kernels, 3D Gaussian Splatting, and vision-language models (Qwen-VL, LLaVA, SAM/SAM 2). I have published first-author papers in IEEE TMM and IEEE Access, demonstrating strong communication skills and the ability to translate complex technical concepts for diverse stakeholders.
Selected Research & Projects
Text-to-Skeleton Cascades for Controllable Complex Human Motion Video Generation
Ashkan Taghipour, Morteza Ghahremani, Zinuo Li, Hamid Laga, Farid Boussaid, Mohammed Bennamoun
SVR-GS: Spatially Variant Regularization for Probabilistic Masks in 3D Gaussian Splatting
Ashkan Taghipour, Vahid Naghshin, Benjamin Southwell, Farid Boussaid, Hamid Laga, Mohammed Bennamoun
This work was conducted during my research internship at Dolby .
Box It to Bind It: Unified Layout Control and Attribute Binding in T2I Diffusion Model
Ashkan Taghipour, Morteza Ghahremani, Mohammed Bennamoun, Aref Miri Rekavandi, Hamid Laga, Farid Boussaid
Ashkan Taghipour, Morteza Ghahremani, Aref Miri Rekavandi, Z Li, Mohammed Bennamoun, Hamid Laga, Farid Boussaid
MineWatchAI — Mining Rehabilitation Monitoring
End-to-end data science application for monitoring vegetation rehabilitation at WA mining sites using Sentinel-2 imagery, vegetation indices (NDVI, SAVI, EVI), and automated compliance reporting.
WealthPathAU — Investment Portfolio Simulator
ASX investment simulator with Monte Carlo projections, historical backtesting, dividend forecasting, and risk-based portfolio allocation serving Australian retail investors.
Experience
Data Scientist / ML Engineer at Novarc Technologies
Apr 2024 - Sep 2025 (Part-Time, Remote) — Vancouver, Canada
- Designed and deployed AI/ML models for industrial video analytics, improving operational throughput by 130% (15 to 35+ FPS) and reducing prediction errors by 20%.
- Developed predictive analytics pipelines for quality control and anomaly detection in manufacturing processes, enabling proactive maintenance and reducing downtime.
- Engineered synthetic data generation frameworks that reduced real-data labelling costs by 60–75%, improving model robustness for industrial robotics applications.
- Applied experimental design and A/B testing methodologies to optimise model performance, achieving 4–8% accuracy improvements on challenging segmentation tasks.
- Collaborated with cross-functional teams (engineering, operations, product) to translate business requirements into technical specifications for data science projects.
Research Intern — Advanced Technology Group at Dolby Laboratories
May 2025 - Sep 2025 — Sydney, Australia
- Developed a Gaussian pruning framework for 3D Gaussian Splatting achieving 5.6× compression while maintaining visual quality for AR/VR rendering (SVR-GS Paper).
- Implemented high-performance computing solutions using custom CUDA kernels, reducing GPU memory footprint by 82% and enabling real-time novel view synthesis on consumer hardware.
- Communicated research findings through presentations and technical documentation to stakeholders across technical and business teams.
AI Researcher (Ph.D. Candidate) at University of Western Australia
Apr 2023 - Present — Perth, Australia
- Published 8 peer-reviewed papers in top-tier journals (IEEE TMM, IEEE Access), demonstrating expertise in statistical analysis, experimental design, and applied analytical modelling.
- Developed and deployed computer vision and NLP models including natural language processing (traditional and LLM-based), image classification, segmentation, and predictive analytics.
- Managed 10TB+ datasets using distributed computing frameworks, building scalable data pipelines with comprehensive documentation and version control.
- Mentored 4 students on research methodology and ML best practices, fostering a culture of collaboration and continuous learning.
Innovation Center Manager — ML & Data Science at MTN Group
Apr 2021 - Apr 2023 — Tehran, Iran
- Built credit scoring and predictive analytics models using PySpark and Databricks, serving 50+ million users and reducing default rates by 10%.
- Designed and maintained data pipelines for large-scale data processing, ensuring data integrity, accessibility, and compliance with security standards.
- Created interactive dashboards and BI solutions serving 15+ stakeholders, reducing report generation time by 40% and enabling data-driven decision making.
- Led continuous improvement initiatives integrating AI solutions into business operations, contributing to 5% revenue increase.
- Managed and influenced stakeholders across technical, operational, and business teams, translating business needs into actionable data science projects.