Curriculum vitae

Shreejal Trivedi

Machine Learning Researcher at Cloudastructure. contact@shreejal.ai · linkedin.com/in/shreejalt · github.com/shreejalt

Experience
8 roles
Patent
1, US
Awards
5

The tags and the pages under a role are this site's: a tag opens the topic, a page the write-up of that work.

Experience

  1. August 2025 – Present
    Remote, India

    Machine Learning Researcher · Cloudastructure Inc.

    • Integrated attribute detection directly into the YOLO architecture, eliminating downstream classifiers.
    • Built a camera mapping framework leveraging SfM, photogrammetry, MCMO tracking, and ground-plane alignment—cutting false alerts by over 50%.
    • Developing an Auto-Annotation pipeline using Prefect, Haystack, and VLMs to enable continuous labeling, online training, and data generation for downstream systems.

    Used YOLO · SfM · photogrammetry · MCMO tracking · ground-plane alignment · Prefect · Haystack · VLMs

    Object detectionTrackingVision and language

  2. May 2023 – August 2025
    Toronto, Canada

    Computer Vision Engineer · Trans-plan

    • Worked with Trans-Plan on 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 actionable data products including traffic density maps, vehicle flow rates, heatmaps, and inbound/outbound counts to support better traffic management.

    Used GDAL · rasterio · shapely · drone footage · georeference · traffic density maps · vehicle flow rates · heatmaps

    Traffic analyticsVideo

  3. June 2023 – July 2025
    Toronto, Canada

    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 plug-and-play edge unit with IP camera and simple management interface.
    • Designed secure 4G data transmission (0.2ms latency) and live traffic dashboard.
    • Containerized stack with recovery, health checks, and data caching for zero data loss.

    Used NVIDIA DeepStream · IP camera · containerized · health checks · data caching

    Traffic analyticsObject detectionTrackingEdge deployment↳ Vehicle Counting UI

  4. Sept 2022 – July 2025
    Toronto, Canada

    Machine Learning Research Assistant · Elderlab, York University

    • Built & deployed a real-time day/night highway analytics stack (detect–track–count) on Highway 401 with Ontario MTO, running on DeepStream + TensorRT edge devices for 24/7 lane, speed, density, and anomaly metrics.
    • Optimised vehicle tracking to <1% daytime and 1.4% nighttime counting error in rain, snow, and fog.
    • Designed unsupervised lane-assignment algorithm via backprojection on ground plane by simulating the oriented bounding boxes around vehicles in real-time.
    • Created an unsupervised domain-adaptation object detection training pipeline using astronomical, nautical, and civil dusk/dawn pseudo-domains, boosting nighttime detection mAP.

    Used DeepStream · TensorRT · oriented bounding boxes · ground plane · astronomical · nautical · civil dusk/dawn · pseudo-domains

    Traffic analyticsObject detectionTrackingDomain adaptation↳ Vehicle Counting UI↳ Unsupervised Domain Adaptation and Semi-Supervised Learning

  5. April 2024 – March 2025 (1 year)
    Toronto, Canada

    Deep Learning Engineer · Eagle Eye Networks

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

    Used pose estimation · action recognition · palm localization · multi-cue gun localization

    Keypoints and poseObject detection

  6. August 2019 – July 2022 (3 years)
    Bangalore, India

    Deep Learning Engineer · UncannyVision (Eagle Eye Networks)

    • Delivered production CV stack (Person Re-ID, detection/tracking, domain-adapted anomaly & background subtraction) and developed an Anchor Search Algorithm for finding dynamic anchors in one-stage object detectors, adding 3% mAP on the in-house baselines.
    • Built a proprietary Site-Specific Training pipeline that raised detector mAP 8% with 500 annotated images and designed a semi-supervised clothing-color classification model which increased the accuracy by 5%.
    • Engineered edge-ready vision modules: 70 FPS Grid-Temporal-Median background subtraction and an unsupervised reconstruction-prediction LSTM trajectory-anomaly detector (98.4% precision, 99.2% recall).

    Used Anchor Search Algorithm · Site-Specific Training pipeline · semi-supervised clothing-color classification · Grid-Temporal-Median background subtraction · unsupervised reconstruction-prediction LSTM trajectory-anomaly detector

    Object detectionRe-identificationAnomaly detectionDomain adaptation↳ Redefining the One-Shot Object Detector for a Two-Class Problem↳ Site-Specific Training Tool↳ Person Retrieval and Re-Identification↳ Low-Compute, High-FPS Background Subtraction↳ Vehicle Trajectory Anomaly Detection

  7. March 2020 – April 2022 (2 years)
    Ahmedabad, India

    Founder and Editor · VisionWizard

    • Mentored in publishing more than 25 deep learning research blogs (140 k reads). Publication: www.medium.com/visionwizard

    Used deep learning research blogs · Medium · VisionWizard

    ↳ The VisionWizard entries on this site

  8. January 2019 – April 2019 (4 months)
    Bangalore, India

    Machine Learning Engineer · UncannyVision (Eagle Eye Networks)

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

    Used pruned · quantized · TensorRT · edge deployment

    PruningQuantizationEdge deployment↳ Network Pruning of Object Detectors↳ INT8 Quantization of Object Detectors

Patents

  • Image Processing and Automatic Learning on Low Complexity Edge Apparatus and Methods of Operation
    One of the inventors.

Education

  • Masters of Science in Computer Science (Thesis)
    York University · Toronto, Canada
    Sept 2022 – July 2025
  • Bachelors in Information and Communication Technology
    Ahmedabad University · Ahmedabad, India
    Aug. 2015 – May 2019

Achievements

  • Innovation York Fellowship · 2024
    CAD 7.5 k, for prototyping real-time analytics system
  • VISTA Scholarship · 2022–24
    CAD 20 k, for scholarly graduate CV research
  • INSPIRE SHE Merit Scholarship
    Government of India
  • Top Writer in Artificial Intelligence on Medium
  • Top 5 in VishwaHack' 2018 Hackathon · 2018
    Smart City

Skills

Languages
JavaPythonC/C++SQL (Postgres)
Frameworks
PyTorchCUDATensorRTTensorflowDeepstreamGStreamer
Developer tools
AWSGCPGitDockerKubernetesFlaskGrafanaPrometheusTimescaleDBKafkaInfluxDBArcGIS Drone2Map
Concepts
Domain AdaptationSSLAnomaly DetectionDiffusion ModelsGANs

Open source

More on GitHub