Tanisha Khurana

Tanisha Khurana

Graduate student in ECE. Specializing in Computer Vision and Machine Learning

Curriculum Vitae

Resume

Download Resume

Contact

Experience

Precision Sustainable Agriculture, NCSU

Graduate Research Assistant

March, 2023 - Present, Raleigh, NC

  • Collaborated with USDA to integrate a camera system with an ML model for mapping crop species, biomass and densities.
  • Working on Domain Adaptation and Multi-task learning for Semantic Segmentation using Deeplabv3+ and Biomass Composition of plant species to assess crop yield and monitor plant growth.
  • Achieved an RMSE of 6.49, a 14.3% decrease from the SOTA model with no additional data and improved real time performance.
  • Engineered a containerized system with integrated RESTful APIs for seamless data visualization and control.
  • Implemented an image classification model for precision farming, accurately differentiating crop species from weeds

Active Robotics and Sensing Lab, NCSU

Research Assistant

March, 2023 - September 2023, Raleigh, NC

  • Performed an extensive literature survey and analysis on region-based and topology preserving edge-based chamber segmentation techniques for identification of Foraminifera species.
  • Generated 2D segmentation masks from synthetic 3D reconstructions in Blender and trained a U-Net segmentation model.

Wobot.ai

Senior Computer Vision Engineer

May,2021 - July,2022 Delhi,India

  • Developed customized Video Analytics and Smart Surveillance solutions for diverse industries including hospitality, food service, and retail, resulting in improved security and operational efficiency.
  • Formulated algorithms for varied tasks including activity recognition, multi-object detection and tracking, pose estimation, motion detection, facial recognition, and person re-identification.
  • Processed RTSP feeds from over 200+ CCTV cameras, enabling advanced monitoring and actionable insights.
  • Scaled ML models in high-throughput and low-latency using TF Serving and triton leading to 50% faster inference time.

  • Implemented Vehicle detection using Yolov5s (License plate detector), Nvidia-LPRnet and Paddle Ocr for license plate detection and detects the wait time if it exceed a specified amount.

  • Activity Recognition for usecases like mopping and shoplifting detection using Slow Fast with Pytorch and DeepSort tracking algorithm.

Intello Labs

Machine Learning Engineer

Jan,2020 - May,2021 Gurgaon,India

  • Led the entire development lifecycle for a real-time AI powered commodity grader Intello Sort utilizing size, color and visual defect analysis.
  • Developed Intello Shelf Eye to monitor shelves and bins and assess the quality of the fruits and vegetables over a Rasberry Pi camera.
  • Accomplished an identification accuracy of 95% and classification accuracy of approximately 90%.
  • Utilized Faster RCNN, Mask-RCNN and SSD for object detection of 20 different fruits with an average size error of ∼1 mm.

Qiggle

Data Scientist

Jan,2019 - Dec,2019 Delhi,India

  • Designed a predictive analytics solution for industrial applications using Anomaly detection and remaining life estimation
  • Detected under-performing and abnormally-behaving assets to save weeks of lost power generation and reduce asset downtime.

Innefu Technologies

Data Science intern

Oct, 2018 - Dec, 2018 Delhi,India

  • Implemented ResNet50 and YOLOv3 for advanced facial recognition and video analytics in collaboration with the Delhi Police, leveraging their law enforcement dataset.

  • Achieved a high True Positive Rate (TPR) exceeding 98% in tasks such as missing persons detection, intrusion detection, crowd management, and security surveillance.

GenElek Technologies

Electronics Engineer intern

Aug,2018 - Oct, 2018 Delhi,India

  • Contributed to the development of cost-effective and lightweight lower limb exoskeletons, pioneering advancements in algorithms for seamless transitions between sitting and standing positions.
  • Led the implementation of actuation and simulation through the integration of Arduino, accelerometer, EMG muscle sensor, and actuators.

Achievements

  • Ranked first in NC Plant Science Initiative hackathon for the most accurate semantic segmentation model with 79% accuracy.
  • Recipient of the GHC 2023 Scolarship and SWE 2023 Scholarship from NC State.
  • Academic Chair for Women in ECE (wiECE) at NC State.
  • Completed a 10 week Leadership Development Program Certification Course at NC State. Earned a leadership certificate by engaging in a coaching relationship and leadership experience, and completing the foundation workshops and a reflection paper.
  • Member of Women in Computer Science (wics), Embedded Machine Learning Club and Quantum Information Club at NC State.

Skills

  • Languages: Python, C/C++, SQL, Bash, Git
  • Developer Tools: MATLAB, AWS, Azure, Google Cloud, Docker
  • Frameworks: OpenCV, Pytorch, Tensorflow, Scikit-Learn, Pandas, Numpy, PIL, Matplotlib, Seaborn,Pytest ROS, Flask

Courses

Digital Imaging Systems, Neural Networks and Deep Learning, Advanced Machine Learning, Computer Vision, Digital Signal Processing, Natural Language Processing, Probability and Random Processes, Detection and Estimation Theory, Pattern Recognition, Cloud Computing, Probabilistic Graphical Models for Signal Processing and Computer Vision