Deepak Phadtare's repositories
DeepakPortfolio
My Work
optimized-etl-fetch
ETL Pipeline: SQS Queue --> PostgreSQL
Harware_Generation_Tool
• Built a tool that flexibly generates hardware to accelerate the evaluation of multiple layers of convolution system with a non-linear function called a ReLU after each convolution • Designed it using digital system design techniques such as control module, data path and FSM • Enhanced the speed by adding parallelism and pipelining • Programmed the tool by C++ for software development and System Verilog to generate the hardware • Tested the tool by using random testbenches for its speed, area, throughput and energy consumption
Machine-Learning
Implemented algorithms like Star cubing, Bayesian Classifier, FP growth, Decision Tree Induction, Logistic Regression, KNN from scratch without using any inbuilt libraries
Object-Detection-By-YOLO
This system, you only look once (YOLO) at an image to predict what objects are present and where they are. By this way we can reframe object detection as a single regression problem, straight from image pixels to bounding box coordinates and class probabilities. Using you only look once (YOLO) at an image to predict what objects are present and where they are.