KARTHIK DATTA YERRAM

Boston, MA 917-***-**** adxile@r.postjobfree.com www.linkedin.com/in/karthik-datta EDUCATION

Columbia University New York, NY

MS in Electrical Engineering, specializing in Machine Learning Dec 2022

• GPA : 3.81/4 Courses: Algorithms of Data Science, Practical Deep Learning and Performance, Machine Learning, Big Data Analytics, Applied Machine Learning, Computer Vision II, Large Scale Stream Processing. Manipal Institute Of Technology, Manipal, India

B.Tech in Electronics and Communication Engineering Jul 2020

• GPA : 8.09/10 Courses: Computer Vision, Motion & Geometry based methods in Computer Vision. EXPERIENCE

Shipin Systems Inc. Boston, Massachusetts

Computer Vision and Deep Learning Engineer Feb 2023 - Jun 2023

• Owned the end-to-end capability of integrating thermal cameras into the Shipin ecosystem. Developed Hotspot detection algorithm using OpenCV to identify potential heat zones in critical areas like Generator, Engine and Oil Water Seperator rooms to alert the relevant personnel.

• Architected an algorithm combining 2D Power Spectral Density analysis with CNN to detect blur/wetness on the camera lens. This innovative approach significantly improved the overall camera health and reduced false detections.

• Consolidated and automated the error analysis pipeline of the object detection model used to analyze false predictions. This led to a reduction of 66.67% of the computation time.

Rivian Automotive, Inc. Palo Alto, California

Computer Vision Intern May 2022 - Aug 2022

• Worked on One Shot NAS (trained Supernet composed of Resnet-50 blocks and used Evolution Algorithm as search strategy) optimized classification backbones for classification tasks upto 65% in terms of GMACs and achieved state of the art accuracy of 87% on CIFAR 10 dataset.

• Optimized Object Detection models using One Shot NAS techniques upto 87% in terms of GFlops compared to baseline YOLOv5-M on the on-board SOC for the use-case of Guard Mode Monitoring feature of the Rivian car and analyzed the results using Weights and Biases. Vimaan Robotics India Private Ltd Banglore, India

Deep Learning Engineer - Intern Aug 2020 - Oct 2020

• Added a feature to Labelme software to propagate annotations of static objects to consecutive frames based on the distance between key points calculated using the SIFT algorithm. Saved 75% of the labelling time.

• Automated propagation of annotations of dynamic objects by calculating optical flow between consecutive frames with the help of Flownet 2.0 and Lucas-Kannade algorithm was performed at the boundaries of every object thereby improving the optical flow. Enhancement resulted in saving over 44% of the time while annotating every frame in a video. Infrrd Private Ltd Banglore, India

Computer Vision Intern Jan 2020 - Jul 2020

• Built a model utilizing Support Vector Machine as the classifier and Histogram of Gradients as the feature extraction algorithm with an accuracy of 87%. Leveraged OpenCV to pre-process the raw images before being fed to the model to verify the output of OCR.

• Developed a Convolutional Neural Network(CNN) model from scratch with an accuracy of 91.3%. The model could predict 120000 images in less than 2 minutes thereby eliminating the role of manual data entry, minimized time, money and labor for the company. Mars Rover Manipal Manipal, India

Electronics and AI Member Oct 2017 - Jul 2019

• Led the Electronics and AI team of "Mars Rover Manipal", ranked "FIRST" all over Asia and "EIGHTH" in the world out of 84 teams in

'University Rover Challenge' organized by The Mars Society at the Mars Desert Research Station (MDRS) Utah, USA.

• Built Mobilenet V2 for the detection of visual marker and implemented the same on the on-board SOC (Jetson TX2).

• Implemented local planning in the use case of object avoidance and autonomous rover traversal using Intel RealSense Depth camera. Analyzed Point Cloud stream provided by the camera using Point Cloud Library. SKILLS

• Python, Pytorch, Tensorflow, WandB, Keras, Numpy, OpenCV, Pandas, Scikit Learn, Spark, BigQuery, GCP, Airflow. PROJECTS

Region Proposal Food Object Detection Sep 2022

• Architected end-to-end pipeline for object detection that circumvents the tedious process of labelling bounding boxes by employing region proposal algorithm and image classifiers trained on Food-101 dataset. Performed exhaustive comparative analysis of MobileNet, VGG16, InceptionNet, EfficientNet and Resnet50 on metrics -Time-to-Accuracy, Memory Utilization, Classification Accuracy. Image Registration using Computer Vision Algorithms Jul 2019

• Stitched images sharing an overlapping region leveraging SIFT algorithm to find key features and Brute Force matcher to find best matches followed by estimating the Homography matrix in order to warp the images. PUBLICATIONS

• K.D. Yerram, I. Shreevastava, "Image Stitching Using Computer Vision Algorithms", Automation Information Technology in Steel Mining Industries -Steel Authority of India & Computer Society of India, Ranchi, India. (ISBN No. ***-**-****-329-7).

Contact this candidate