Portfolio of
ISHWAR SONI
Computer Vision & Motion Processing Engineer
Specializing in human motion synthesis and processing large-scale datasets (AMASS). I build robust preprocessing pipelines and intelligent systems that bridge research and production.
Engineering Mindset
I enjoy solving confusing technical problems more than building demo projects. My work revolves around making messy systems reliable, whether it's normalizing AMASS motion datasets or deploying prediction models to production.
/Engineering mindset: I prioritize reliability and reproducibility.
/Debugging complex systems: I dive deep into coordinate transforms and data inconsistencies.
/Working with messy data: I build robust pipelines to handle real-world edge cases.
/Shipping over showing off: I value deployed, working systems over theoretical perfection.
Experience
Semantic Labs
Worked on large-scale AMASS motion datasets and BVH ↔ SMPL-H preprocessing pipelines.
- Built and optimized BVH ↔ SMPL-H preprocessing pipelines for large-scale datasets.
- Fixed critical coordinate system normalization issues affecting downstream models.
- Resolved orientation flips and scaling bugs in motion data.
- Implemented motion stabilization and smoothing algorithms.
- Optimized data pipeline performance for faster processing.
Featured Projects
A collection of technical case studies focusing on system architecture, data pipelines, and scalable infrastructure.
BVH-to-SMPLH Motion Processing Pipeline
Technologies
Computer vision and motion processing pipeline for BVH to SMPL-H human motion reconstruction.
The Challenge
BVH motion files contained coordinate-system inconsistencies, orientation flips, scaling errors, and schema mismatches that produced unstable SMPL-H reconstructions.
The Solution
Designed transformation and preprocessing pipelines to resolve coordinate alignment issues, improve motion stability, and validate reconstruction quality through visualization workflows.
Project Impact: Demonstrates expertise in human motion modeling, motion processing, skeletal transformations, and large-scale computer vision data pipelines.
Interactive Computer Vision system for real-time gesture-controlled visual effects.
The Challenge
Creating stable and responsive AR effects that accurately follow hand movements in real time while maintaining smooth visual rendering and low latency.
The Solution
Implemented MediaPipe-based hand tracking and a modular rendering pipeline to generate gesture-controlled Rasengan and Chidori-inspired visual effects synchronized with user movements.
Project Impact: Demonstrates real-time computer vision, gesture recognition, and AR/VFX engineering through an interactive hand-tracking experience.
Automated machine learning data preprocessing framework for robust and consistent dataset preparation.
The Challenge
Real-world datasets often contain missing values, outliers, skewed distributions, and redundant features, making manual preprocessing time-consuming and error-prone.
The Solution
Developed a scalable data cleaning framework that automates preprocessing workflows, applies safety checks against data leakage, and handles common data quality issues across diverse datasets.
Project Impact: Validated on 50+ datasets, demonstrating reliable preprocessing across diverse data distributions while reducing repetitive manual cleaning tasks.
Technical Skills
Computer Vision & Motion
- Human Motion Analysis
- Pose Estimation
- SMPL
- SMPL-H
- BVH
Machine Learning & Data
- NumPy
- Pandas
- Scikit-learn
- Model Evaluation
- Data Engineering
Programming & Tools
- Python
- Git
- GitHub
- Jupyter Notebook
- Docker
- FastAPI