Shankar Kumar's repositories
platooning-demonstrator
A Platooning demonstrator consisting of a convoy of 1/10 scale RC cars. Deep Learning and Computer Vision concepts utilized and system implemented on the Jetson Nano.
Realtime-voice-cloning-as-a-microservice
SV2TTS as a Microservice (FastAPI endpoint)
MAFAT-RADAR-Challenge
The MAFAT challenge, by the Israeli Department of Defense. Deep Learning based approach to classify radar signatures of humans and animals.
accelerated-incremental-learning-with-support-vector-machines
Accelerated implementation of the SVM algorithm on an embedded SoC with changes to the algorithm structure to accommodate learning from large volumes of data in an incremental manner with approximations to the computations.
Rommelag-Defect-Identification-Challenge
Defect Identification Challenge to find an efficient and exact process for error detection in the production line using Computer Vision, Deep Learning or Classical Machine Learning based techniques.
Pepperl-Fuchs-Lidar-Classification-Challenge
To increase transparency and efficiency in intralogistics, we are looking for a solution to classify moving boxes in a production area. By using the latest 2D-Lidar technologies powered by Pepperl+Fuchs, you will generate valuable insights. Get creative and pull out anything of your machine learning toolbox!
pose-estimation-with-high-resolution-neural-networks
Simpler reimplementation of Pose Estimation with HRNets using a Top Down Approach.
activity_recognition
Activity Recognition using a Linear Classifier based on Accelerometer Data
Birmingham_City_Uni_AI_Hack
The aim of the hack was to deliver AI-powered solutions which could inform medical research, policymaking and knowledge dissemination on the pandemic, as well as raising awareness among the public, to investigate the hidden knowledge inside a World Health Organisation dataset summarising Covid-19 daily-case and death-rates from each reporting country.
FAQ-Generation-and-SQuaD-Style-QA-Answering-System
An FAQ dataset generation system that considers noun phrases and verb phrases as the answers, and a QA SQuaD Style dataset generation system that uses this data to train a BERT based QA system.
home-work-location-recognizer
Predicting the likely Home and Work locations of a person based on Clustering and GPS coordinates
learning-to-smell
Molecules are the actual building blocks of all fragrances. We take these molecular compounds as an input, parse them through, and predict what multitude of fragrances they contain out of 100+ different ones.
lincoln-poinsettia-challenge
The Lincoln Agri-Robotics Poinsettia Challenge is a machine learning and computer vision competition to devise intelligent ways to identify features that contribute to rating the "best" poinsettia
recognize-named-entities-on-twitter-with-LSTMs
Recognize named entities on Twitter with LSTMs
Similarity-Measures-for-Music-Loop-Projects
Similarity Measures for Music Loop Projects in the Music Maker JAM Application
transfer-learning-with-Inception-v3
Showcasing Transfer Learning with the Inception V3 on the Stanford 40 Actions Dataset
AES-Encryption
AES Encryption
artistassistapp
ArtistAssistApp - the web app to paint better with ease. Tools for realistic color mixing based on real paints, tonal value drawing, simplified sketching, and more. Perfect for artists of all levels and styles, whether you prefer watercolor or oil, realism or loose techniques, landscapes or portraits. Try it now!
Azure-MLOPs-Helper
Custom Helper Class that encapsulates common functions for llm deployments on VMs.
Game-of-Life
Game of Life
Gaussian-Processes-for-Regression
Gaussian Processes for Regression
Simulated-Annealing-for-Travelling-Salesman-Problem
Simulated Annealing for the Travelling Salesman Problem
tag-prediction-for-StackOverflow-questions
Predict tags on StackOverflow with linear models