Name: MANISH KUMAR PANDEY
Type: User
Bio: Co-Founder
@FreeDoctr
, Research Engineer Stealth #GraphML, #GeometricDL, #3DComputerVision, #DiffusionModels, #Generative AI #ML ,#RL, #LLM, #BioInformatic
Twitter: Manish_GenAI
Location: New Delhi
Blog: https://www.linkedin.com/in/manish-genai/
MANISH KUMAR PANDEY's Projects
Palmprint ROI extraction under unconstrained pose variations and implementation of a Deep Learning based system for contactless palmprint verification.
All graph/GNN papers accepted at the International Conference on Machine Learning (ICML) 2024.
Team Reporting App Flask Backend
Datasets for Data Analytics Tasks
Deep-Learning-Based Approach to Anomaly Detection Techniques for Large Acoustic Data in Machine Operation.Developed a deep leaning algorithm which detects anomaly in acoustic sensor data with approx. 90% accuracy. ļ§ Implemented the different machine/deep learning algorithms like SVM, KNN, K-means, CNN, Delayed LSTM, Conv LSTM and different Beamforming algorithms such as delay and sum beamforming, linear constrained minimum variance beamformer etc. and analyzed their limitations ļ§ Formulated the Sound source localization algorithms like MUSIC algorithm (Multiple Signal Classification), TDOA and Steered response and currently working on the optimization of it using GAN-LSTM
ā¢ Developed a Deep Learning-based Covid-19 Time Series Prediction. ā¢ Used Deep Learning and Statistical approaches to capture the patterns and trends of varying events related to infectious diseases. ā¢ Implemented ARIMA,HWAAS Models for exploiting linear dependencies in observations and time series forecasting for univariable data. ā¢ Explored RNN, LSTM Neural Network to find temporal correlations in time series prediction.
DevOps Workshop
StamLit
EEML Code Implementation for the paper "ADAPTIVE UNIVERSAL GENERALIZED PAGERANK GRAPH NEURAL NETWORK"
Email Automation Pipeline
This documentation is like a quick snapshot of my project in the data field, showing off my skills and know-how in this area.
The Receipt Processor is a web service built with Go that processes receipts and calculates points based on predefined rules. It provides a simple and efficient way to process receipts and determine the points earned for each receipt.
Foundations
This playlab encompasses a multitude of projects crafted through the utilization of Large Language Models, showcasing the versatility and impact of these models across various applications.
Repo for the overall practice
API WebScrapper
Real Estate Valuation Model: Predict property prices with our ML model. Developed using Linear Regression, it estimates values based on house age, MRT distance, and store count. Includes code, documentation, and instructions for data preprocessing, model development, evaluation, deployment, and testing. Regularly updated for accuracy.
Python-based Implementation for "Respiratory Scans-based COVID-19 Detection using Multi-Modal Multi-Task Learning Framework"