Mahendran Narayanan's repositories
SENetV2-Aggregated-dense-layer-for-channelwise-and-global-representations
Code for paper SENetV2 Aggregated dense layer for channelwise and global representations
Academics-module
College project to develop Registration module and Graduation module for IIT Tirupati. These modules are developed during the coursework as Software prototype. This serves as base for full scale development of ERP based website in IIT Tirupati.
Fitness-tutor
We present Fitness tutor, an application for maintaining correct posture during workout exercises or doing yoga. Fitness tutor which can guide the workouts using pose estimation. We use posenet model in tensorflow with p5js for developing skeleton. Fitness tutor is an application of pose estimation model in bringing a realtime teaching experience in fitness.
GPT-finetuning
Code for finetuning GPT models and its variants for text data
Medullo
Medullo is a web application that provides a platform for users to share and search for resources. The core idea of this application is to provide users with a medium to share their resources thereby enabling them access to resources, instead of having to buy or download multiple copies of the same resource. This application could well be used among students to share resources like class lecture notes or lecture videos or solution to problems.This can also help students connect with other students who are pursuing similar courses of interest and engage in a communal process of passing down study materials. It can also help users in general, to find people with similar interests and share the things they love. This encourages the habit of reducing and reusing which could in turn contribute to the global process of waste reduction in the long term. Once a user wants a specific resource, a resource for that request is raised. Once the request is approved by the resource owner, contact of the resource owner is shown. This feature provides for the privacy of resource owner and eliminates spamming. The upload feature enables for sharing of resources with the global community.
Memory-Visualization-for-Deep-Neural-networks
A tool for analysis of memory utilized for developing and training deep learning models. Our tool results in visual utilization of memory concurrently. Various parameters affecting the memory utilization can be analysed while training with this tool. This tool also helps in knowing better idea of processes or models which consumes more memory.
SD-SAM-for-Background
Using Stable diffusion (SD) model and Segment Anything model from Meta (SAM) for removing the background and add newer background for the image. Created with a Multi model pipeline.
Transportation-booking
A simple online booking form for Faculty & Staff, Student and Guest for availing the transport both for official and personal use. Used to connect two campuses of IIT Tirupati.
Django-forms
Create a form in Django and update it in MySQL backend
GPT-unlearning
Testing a simple example to understand Machine unlearning concept with GPT. GPT is made to forget Alphabets
image-processing
This repository contains all kinds of simple image processing tasks
ivy
The Unified AI Framework
LeNet-Tensorflow-Pytorch
Create LeNet model including Tensorflow and Pytorch packages
Mobilenet-Tensorflow-Pytorch
Code for Mobilenet and its variants. Code is in both Tensorflow and Pytorch
Neural-network-maker
The code creates deep learning CNN model automatically with very basic parameters. It includes popular modules as well while creating.
pytorch-1
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Resnet-Tensorflow-Pytorch
Code for resnet and its depth variants. Code is in both Tensorflow and Pytorch
SAENet-Tensorflow-Pytorch
Code for Squeeze aggregated excitation resnet and its depth variants. Code is in both Tensorflow and Pytorch
SENet-Tensorflow-Pytorch
Code for Squeeze and excitation residual network and its depth variants. Code is in both Tensorflow and Pytorch
Tensorflow-Pytorch-model-converter
This project enables the model conversion between languages Tensorflow and Pytorch. Model to be converted is sent as input and it results the model in other language.
VGG-Tensorflow-Pytorch
Code contains VGG model in Tensorflow and pytorch
vision
Datasets, Transforms and Models specific to Computer Vision
vision-transformer
Vision transformer Python
Webscraper
Scraping the data from web