TAANYA GUPTA

adjkr2@r.postjobfree.com 657-***-**** Riverside, CA- 92507 Linkedin Github

EDUCATION

Master’s in Science University of California, Riverside, CA – 3.53/4.0 Sept 2018- June 2020 Electrical and Electronics Engineering

Coursework: Pattern Recognition, Advanced Computer Vision, Machine Learning, Data Mining Techniques, Stochastic Processes, Data Center Architecture, Intelligent Transportation Systems, Deep Learning Bachelor’s in Technology Punjab Engineering College, Chandigarh, India April 2014 – March 2018 Electronics and Communication Engineering

Career Objective: Computer vision and Deep learning enthusiast looking for full time opportunities June 2020 TECHNICAL SKILLS

Languages and Frameworks Python, C++, C, MATLAB, Java, R, MySQL, TensorFlow, PyTorch, Keras, Caffe, Hadoop, Spark Libraries, tools, OS OpenCV, SciPy, Matplotlib, scikit- learn, pandas, NumPy, Windows, Linux, AWS Machine Learning EXPERIENCE

Data Analyst, Capital One, Irvine, CA December 2020- Present AI/Machine learning Intern, Happiibook, Los Angeles, CA October 2020- November 2020

• Development of a recommendation system for a new social networking application Machine Learning Engineer, TensorIot, Irvine, CA April 2020- June 2020

• Development of Machine Learning algorithms incorporating AWS Cloud services such as Sagemaker and make use of AWS trainings in Machine Learning, Technical professional, Business Professional

• Optimizing and deploying machine learning models on edge devices, training models, deploying projects in client environments Graduate Thesis, University of California, Riverside, CA (Bir Bhanu, Distinguished Professor) June 2019- June 2020 ICPR paper accepted-Wildfire Smoke Detection using Computer vision in Deep Learning – Research and Implementation

• Researched and developed a smoke detection pipeline which conjoins the spatial and temporal features of wildfire smoke using video object segmentation and optical dense flow in a fully convolutional network

• Mitigates the paucity of labeled data for wildfire smoke videos by integrating Mask RCNN image segmentation technique

• Pre- processed complex image data using physical features of haze which improves the IOU for the predicted mask Computer Vision & Image Processing Intern, Quantum Design, San Diego, CA June 2019- Sept 2019

• Designed algorithms for autofocus with various sharpness metrics for the image enhancement with respect to correcting astigmation and defocus for SEM images

• Improve the image quality of SEM images using various image processing and machine learning techniques

• The automation reduced the time taken to manually adjust by almost 20% Data Analysis & Visualization Intern, Microsoft Innovation Center, Nepal July 2018- Aug 2018

• Data processing and analysis using R to perform robust evaluations for existing and new clients such as UNICEF

• Redesigned and devised Power BI data visualization dashboards for clients and imported it to web application.

• Programmed and debugged web-based application using .NET and prototype of the applications using WireFrame

• Conducted several KPIs to utilize the various services PROJECTS

Comparison Sift, SURF, and ORB

• Analyze three main feature matching methods, SIFT, SURF, and ORB, and their suitability, or lack thereof, to real-time applications.

• We will evaluate their performance in terms of computational cost and robustness on the graffiti dataset Face Swap Using Generative Adversarial Networks

• Employ deep learning techniques, such as object detection, image segmentation and generative adversarial networks to swap the faces of two people in two videos.

Individual Recognition in a video using GAIT

• An alternative to the current biometric techniques to identify a human. It incorporates individual recognition by creating energy images through Silhouette images using various computer vision techniques Data mining techniques for Crime Gang

• Identify a specific gang and the weapons used by criminals used in the area of Los Angeles using the three data mining techniques to extract the information from Kaggle data.

License Plate Recognition

• Designed an intelligent recognition method using optical character recognition in OpenCV using the basic concepts on machine learning.

Contact this candidate