Shreejal Trivedi

Machine Learning Researcher

Cloudastructure

• Integrated attribute detection directly into the YOLO architecture, removing downstream classifiers and running 4x faster.
• Built a camera mapping framework using SfM, photogrammetry, MCMO tracking and ground-plane alignment, cutting false alerts by over 50%.
• Developing an auto-annotation pipeline with Prefect, Haystack and VLMs for continuous labeling, online training and data generation.

Computer Vision Engineer

Trans-plan

• Advanced traffic analytics using drones and infrastructure cameras for real-time highway monitoring.
• Improved Python workflows to process and georeference dynamic drone footage using GDAL, rasterio and shapely.
• Delivered traffic density maps, vehicle flow rates, heatmaps and inbound/outbound counts.

Machine Learning Engineer

Ontario Ministry of Transportation (MTO)

• Built real-time vehicle counting with lane-level classification and adaptive clustering.
• Ported to NVIDIA DeepStream with custom plugins for counting, tracking and messaging.
• Deployed a plug-and-play edge unit with IP camera and a simple management interface.
• Designed secure 4G data transmission and a live traffic dashboard.
• Containerized the stack with recovery, health checks and data caching for zero data loss.

Machine Learning Research Assistant

Elderlab, York University

• Built and deployed a real-time day/night highway analytics stack (detect, track, count) on Highway 401 with Ontario MTO, on DeepStream + TensorRT edge devices.
• Optimised vehicle tracking to under 1% daytime and 1.4% nighttime counting error in rain, snow and fog.
• Designed an unsupervised lane-assignment algorithm by backprojection onto the ground plane.
• Created an unsupervised domain-adaptation training pipeline using astronomical, nautical and civil dusk/dawn pseudo-domains, raising nighttime detection mAP.

Graduate Teaching Assistant

Lassonde School of Engineering, York University

• EECS1730 Building Interactive Systems (Jan – Apr 2025)
• EECS2101 Fundamentals of Data Structures (Jan – May 2024)
• EECS1015 Introduction to Computer Programming (Sep – Dec 2023)
• EECS1720 Building Interactive Systems (Jan – Apr 2023)
• EECS1015 Introduction to Computer Programming (Sep – Dec 2022)

Deep Learning Engineer

Eagle Eye Networks

• Built a multi-cue gun localization algorithm (pose estimation, action recognition, palm localization); recall up 13%.

Deep Learning Engineer

UncannyVision / Eagle Eye Networks

• Delivered a production CV stack (person Re-ID, detection and tracking, domain-adapted anomaly detection, background subtraction) and an Anchor Search algorithm adding 3% mAP on in-house baselines.
• Built a site-specific training pipeline that raised detector mAP 8% with 500 annotated images, and a semi-supervised clothing-colour classifier that raised accuracy 5%.
• {'Engineered edge-ready modules': '70 FPS grid-temporal-median background subtraction and an unsupervised LSTM trajectory-anomaly detector (98.4% precision, 99.2% recall).'}

Projects from this role:

• Redefining the One-Shot Object Detector for a Two-Class Problem
• Low-Compute, High-FPS Background Subtraction
• Site-Specific Training Tool
• Vehicle Trajectory Anomaly Detection
• Person Retrieval and Re-Identification
• Unsupervised Domain Adaptation and Semi-Supervised Learning

Founder and Editor

VisionWizard

• Mentored the publishing of more than 25 deep-learning research blogs (140k reads).

Machine Learning Engineer Intern

UncannyVision / Eagle Eye Networks

• Pruned and quantized models for edge deployment, giving 3x speed and 2.5x smaller footprints.
• Implemented TensorRT from scratch for a further 1.2% mAP and 1.5x throughput.

Projects from this role:

• Network Pruning of Object Detectors
• INT8 Quantization of Object Detectors