Amanmeet Garg Email & Phone Number

United States

Gurgaon, Haryana, IN

Seattle, Washington, US

Emeryville, CA, US

- Developed a video scene content detection model for automated video summary creation.- Tech lead for video Automated Content Recognition (ACR) module.- Large scale ( 150 TB+ per month ) pipeline for video processing.- Podcast summarization, video summarization, Graph neural networks, image compression and retrieval.

Los Angeles, CA, US

• Developed a semantic segmentation CNN model for human body parts in images.
• Developed a general Object tracker model for videos with siamese network.
• Developed a scene change detector algorithm for video scene changes.
• Deployed a video analysis pipeline with neural network models, pre and post processing algorithms.
• Tools: Pytorch, python, iOS coremltools, openCV.

Oakland, California, US

Built 2D object segmentation algorithm for small object detection.Built stereo matching based 3D object location algorithm for robot guidance.Built stereo calibration system for 3D localization in 2 camera setup.C++, OpenCV, ROS ( message passing ), UML software architecture diagrams.

• Developed and deployed a computer vision algorithm for video facial acne analysis deployed in iOS application.
• Developed a system and manually labelled image data for facial acne analysis.
• Developed a first-of-its-kind small object detection, convolutional neural network model pipeline for skin acne detection.
• Deployed a trained Keras deep neural network model in iOS application used by 5000+ daily users.
• Deployed trained model on AWS Lambda cloud for delivery on web applications.
• Tools: Tensorflow, python, numpy, Keras, iOS Coremltools, AWS, c++, openCV, JSON.

Vancouver, British Columbia, CA

1. Developed two Neural networks; 1) Convolutional Neural Network (CNN) architecture and 2) a Recurrent Neural Network (RNN) architecture for time series signal segmentation for automated human activity recognition with 94% accuracy and 92% Sensitivity.2. Manually curated data sets from company hardware for ML systems.3. Implemented a sensor fusion filter.

Burnaby, BC, CA

1. Developed a method for Topology Data Analysis of brain geometry with ~80% classification accuracy in predicting Parkinson's disease.2. Developed a Shape Topology method and Surface displacement shape feature for shape analysis of subcortical brain anatomy.3. Developed Ground truth MRI segmentation atlas for improvement in segmentation from 84% to 92%.4..

Burnaby, BC, CA

Human Anatomy; Human physiology; Active health; Exercise and work physiology, Digital image processing.

Burnaby, BC, CA

1. Developed a Physiological model for cardiovascular and postural control system interaction.2. Statistical validation a Wavelet transform coherence pipeline for analysis of biomedical signals.3. Conducted physiology data acquisition experiments with human participants.

Completed two healthcare analytics projects crucial to the core product of the company.1. Developed a clinical reporting and analytics visualization system.2. Developed a image analysis system for mobile health application.3. Developed and conducted research to transform of in house developed algorithms for image analysis into C++.Tools: Matlab, JSON reader.

Vancouver, British Columbia, CA

1. Manually curated a dataset for skin mole images with malignant, benign lesions via web scraping.2. Developed Deep Convolutional Neural Network (CNN) for prediction of Skin Cancer from dermoscopy images.Tools: Torch, Git, Scrum, Python.

Vancouver, B.C., CA

Developed methods for analysis of MRI data in Post Traumatic Stress Disorder and Concussion patients.

Vancouver, B.C., CA

1. Conducted a study and and analysis of posture data in hockey players suffering from concussion.2. Integrated a Cardio-postural model into clinical assessment methodology.

Vancouver, BC, CA

New York, New York, US

Developed an application for the automation of the recurrent payments.Technology: Java, GUI in Java.

Kharagpur, West Bengal, IN

Did a study on the feasibility of the use of ultrasound imaging for the near skin imaging of blood vessels.

At CSIO I was a part of the team responsible for the development of imaging module for radiation therapy and developed an application in VC++

Doctor Of Philosophy (Ph.D.), Medical Image Analysis

Master Of Science (M.Sc.), Biomedical Physiology And Kinesiology

Bachelor Of Engineering (B.Eng), Electrical Engineering

10Th, High School