Gokul Adethya's Projects
Rendering huge 3D Model scene chunk by chunk efficiently using Oct Tree
Fine-tuning pre-trained ResNet50 model to classify the brain scan different types of tumor
Solving time series problem with neural network
Fine-tuning Resnet18 model to predict if the patient has Pneumonia (COVID19) with the help of x-ray
Tuning different Pytorch/Tensorflow pre-trained models like ResNet50 , Wide ResNet_50.2 , VGG16 and a custom CNN model to classify a dog image among 120 breeds
Training Siamese model to create a one-shot neural network model using Face-Net and MTCNN as the backbone. Yolov5 and OpenCV is used alongside to classify faces in the webcam feed
Official implementation of the paper "Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection"
Server code for keres game hosted in heroku
Solutions of L&T EduTech Hackathon by our team - Bharatfly_Coders
Classifying IMDB reviews using LSTM with pre-trained GloVe embeddings
Comparing perfomance of vgg16 against alexnet in neural style transfer
Easy way to convert scanned documents into editable text documents, classifying key-value pairs and annotating them. MTX - HackOlympics 2.0 Shaastra - 2022
We are trying to compare train losses, training time, and model convergence for RNN , LSTM, and GRU models using sin(x) graph in CPU and GPU
š [CVPRw] SAM-PM: Enhancing Video Camouflaged Object Detection using Spatio-Temporal Attention, Muhammad Nawfal Meeran, Gokul Adethya T, Bhanu Pratyush Mantha
Training U-Net and Mask R-CNN models for instance segmentation in a neuronal cell
offical seesawfacenet pytorch implement, https://arxiv.org/abs/1908.09124
Experimentic with Siamese algorthim
This project integrates T5 like LLM summarization models with Torch Serve, enhanced by extractive methods like LexRank and TextRank. It also includes keyword generation and implements query-based content filtering using cosine similarity, offering a comprehensive solution for article summarization and analysis.
Made an app that classfies if there is a crack or not in the image and tags the place and time which will be shown in a common Dashbaord along with previously taken images
A simple TODO list C program
EmailedIn - The Ultimate LinkedIn Profile Finder
Yet Another Python Compiler