Giridhar NR

I am currently enrolled in the Master's program in Computer Science with a focus on Artificial Intelligence at the University of Southern California.

I am working as Postgraduate Researcher at Yale University under Dr. Daniella Meeker and Dr. Hua Xu

Previously I worked as

Email  /  CV  /  Research Exp  /  Linkedin  /  GitHub

Research Interests: Computer Vision, Generative AI, Machine Learning, Deep Learning, Natural Language Processing, Reinforcement Learning and Robotics.

profile photo

Aug '23  

 Joined Yale University as Postgraduate Researcher under Dr. Daniella Meeker and Dr. Hua Xu

May '23  

 Joined Amazon Science lab - Buyer Risk Prevention Applied Scientist Intern under Dr. Dmitry Pavlov.

Apr '22  

 Joined Neuro Image Computing Research Lab under Dr. Yonggang Shi.

Mar '22  

 Joined Dr. Daniella Meeker's team as a Machine Learning Student Researcher

Jan '22  

 Started MS in Computer Science - AI at University of Southern California.

Oct '21  

 Recevied Star Startup Award from Ramaiah Evolute.

Aug '21  

 Raised funding for AIvolved from Ministry of Electronics & Information Technology and Ramaiah Evolute.

Nov '19  

 Co-founded AIvolved Technologies Pvt Ltd as CTO and Director.

Jul '19  

 Presented IEEE paper in ICCCNT-IIT Kanpur.

Nov '18  

 Published paper in Springer; SAI-Goa.

Oct '18  

 Presented IEEE paper in I4C-Bangalore.

Aug '18  

 Joined Philips R&D-Bangalore as a Machine Learning Research Engineer.

July '18  

 Presented IEEE paper in SAI Computing Conference London-UK.

July '18  

 Co-hosted ERCICA-2018 International Conference - Bangalore.

Apr '18  

 Represented South India in DRDO DRUSE Robotics Competition.

Apr '18  

 Obtained Copyright from Govt. of India for final year thesis.

Dec '16  

 Placed 2nd among 850 teams in Nokia Innovation Day - Bangalore.

May '16  

 Joined Centre for Robotics Research - NMIT.
University of Southern California; Los Angeles, USA.
Master of Science, Computer Science
January 2022 - December 2023

  • Research Focus - Multi site MRI Data Harmonization for Alzheimer’s Disease Analysis. AI for Clinical Empathy
  • Coursework - Machine Learning, Deep Learning, Digital Image Processing, Natural Language Processing, Reinforcement Learning and Advanced Computer Vision
Nitte Meenakshi Institute of Technology; Bengaluru, India.
Bachelor of Engineering, Computer Science
August 2014 - July 2018

  • Research Work - Modelling Soft Transitions in a Video; Robotics - Unmanned Ground Vehicle; Face and Facial Expression Detection.
  • Coursework - Digital Image Processing, Machine Learning, Robotics
Yale University - Biomedical Informatics & Data Science; New Haven, CT
Postgraduate Researcher
August 2023 - Present

  • Research Focus - Generative AI, Language Models
  • Designed and developed a language model that achieved prediction of speakers' lexical attributes. Clean and preprocess authentic clinical oncology encounter data for model training. Evaluate and refine model's performance to ensure its resilience and reliability across diverse applications. Boosted performance and accuracy of speaker recognition from 0.73 to 0.79 by fine-tuning BERT through architectural modifications and precise tuning.

Amazon Science - Buyer Abuse Prevention; San Diego, CA
Applied Scientist Intern
May 2023 - August 2023

  • Research Focus - Continual learning
  • Built an efficient abuse prevention system using Memory Relay and Regularization based Continual Learning. Developed an attention based Continual Learning, achieving a 2% less forgetting over SOTA methods. Enhanced XGBoost’s performance by 1% AUC on incorporating memory-replay continual learning
Philips Innovation Campus; Bengaluru, India
Machine Learning Research Engineer
August 2018 - December 2021

  • Research Focus - Computer Vision, Deep Learning, Ultrasound Medical Imaging
  • wAssist AI - Optimized a Deep Learning Classifier to run on a Intel Celeron Processor.
  • Fetal Heart - Developed a real-time pose estimation model to track systole and diastole phases of a fetal heart. Created deep learning architectures for high accuracy semantic segmentation of anatomies of fetal heart. Researched and developed a reinforcement learning agent to acquire and learn behavior of key planes in a fetal heart
AIvolved Technologies Pvt Ltd; Bengaluru, India
Co-founder & CTO
November 2019 - December 2021

University of Southern California; Los Angeles, USA.
Deep Learning Student Researcher
March 2022 - May 2023

    Dr. Shi's Lab
  • Research Focus - Generative AI
  • Objective - Harmonize the data points in MRI modality for a better classification of Alzheimer's using Adverserial Networds and Transformers.
    • Dr. Meeker's Lab
  • Research Focus - Language Models
  • Objective - Devising an automatic speech and language processing pipeline to diarize and recognize using RNN-Transducers and Attention based model.
Nitte Meenakshi Institute of Technology; Bengaluru, India.
Research Assistant
May 2018 - October 2019

  • Designed a Mathematical model to model Soft Transitions of video and obtained copyright: SW-14707/2017 from Govt. of India
  • Led a Multi-disciplinary team of students at Centre of Robotics Research, to DRDO - DRUSE (DRDO Robotics and Unmanned Systems Exposition
  • Designed and Developed a Human Tracking Mobile Robot for Defense Application
  • Designed a Face Detection and Facial Expression Recognition algorithm using Digital Image Processing techniques. Paper was presented at SAI Computing Conference 2018- London, UK.
  • Worked on funded projects from Department of Science and Technology (DST), Govt. of India and Vision Group of Science and Technology (VGST), Govt. of Karnataka

  • Registeration Number: SW-14707/2017
    • Dated: 04/10/2018
    Title: Modelling of Transitions in Video using Textures

    Filed

  • Application Number: PCT/EP2021/081925
    • Title: Methods for Guided 3D ultrasound acquisition using Spatio-temporal image correlation

  • Application Number: PCT/EP2021/080229
    • Title: Automatic Intelligent Visualization and Interaction using Real time View Plane classification and Pose Estimation

  • Application Number: 21184713.2
    • Title: Automating Localization and Estimation of Heartbeat in First Trimester Ultrasound Scans

  • Application Number: 2020ID01190
    • Title: AI Based approach to improve ultrasound image quality

  • Individual Award (Philips): Bringing wAssist-AI from research prototype to product in a record time
  • Individual Award (Philips): Take ownership to deliver fast’ to boost the accuracy of algorithm from 69% to 84%
  • DRDO: DRUSE Design and Development of Human Tracking Mobile Robot for Defense Application. Top 10 among 15000 teams to represent South India
  • 2nd position, Modern Traffic Management System, among 850 teams in Nokia Innovation Day

 e-FTUS: an Early First Trimester Ultrasound Scan assistance
Soumabha Bhowmick, Giridhar NR, Celine Firtion, Karthik Krishnan, Subhendu Seth, Pallavi Vajinepalli
Philips Research Global - OCUPAI 2020
Abstract: Automatic Detection of Intrauterine pregnancy or ectopic along with clincial findings such as Gestational Sac diameter etc.
Real Time Deep Pose Estimation in Ultrasound
Karthik Krishnan, Giridhar NR, Celine Firtion, Pallavi Vajinepalli
Philips Research Global - OCUPAI 2020
Abstract: High quality, high frame rate localization of 12 anatomical keypoints in the Fetal Heart 4 Chamber view and 2 anatomical keypoints in the Femur.
Modelling Fade Transition in a video using Texture Methods
Giridhar NR, Aniketh Manjunath, Jharna Majumdar
ICCCMLA-Goa, 2020   Springer-Singapore
Paper
Abstract: A polynomial is modelled to detect soft transitions in a video when converted to texture domain.
Keywords: Gray level co-occurence Matrix, Laws Texture, Statistical Method, Gradient Descent, Fade-Dissolve-Wipe Transition.
Autonomous Mobile Robot Navigation on Identifying Road Signs using ANN
Giridhar NR, Gagan PE, Jharna Majumdar
IEEE - IIT Kanpur, 2019
Paper
Abstract: Features from corner detectors are extracted and trained on supervised learning techniques to identify road signs. MSER is used to detect ROI and LQR controller is developed for robot navigation.
Keywords: Corner Detectors- SIFT, SURF, ORB. Support Vector Machines, Logistic Regression, ANN, Maximally Stable Extremal Regions (MSER), Linear Quadratic Regulator (LQR).
Optical Flow for Detection of Transitions in Video, Face and Facial Expression
Aniketh Manjunath, Giridhar NR, Jharna Majumdar
IEEE - SAI Computing Conference, London-United Kingdom, 2018
Paper
Abstract: Optical flow for determining shot transitions (Descriptors: SIFT) in a video sequence and human-face expression detection (Descriptors: HOG).
Keywords: Scale Invariant Feature Transform, Hough Transform, Optical flow, Skin Segmentation, Histogram of Gradient.
Human Tracking by a Mobile Robot in Low Illumination Environment
Sudip Chandra Gupta, Giridhar NR, Jharna Majumdar
IEEE, 2018
Paper
Abstract: A mobile robot is used for tracking in low illumination. A robust Control system is required for the robot to efficiently follow the human using the computer vision (adaptive low illumination) algorithm. Linear Quadratic Integral (LQI) has been implemented on the system for velocity control.
Keywords: Low Illumination, Tracking, Particle filter, SSMR, Control System, LQI