NISHI KANT MANDAL's Projects
AI Machine learning Deep learning Computer vision NLP Projects
Additional materials to "Design of cooperative communication system based on CubeSat satellites" (master thesis) .
This repository explores diffusion models for medical image data augmentation, crucial for enhancing machine learning model robustness in medical imaging. The four implemented methods include training without augmentation, utilizing keras ImageDataGenerator, employing a DCGAN for image generation, and leveraging DDPM for augmenting images.
The project's name is "EvoPictura: AI-Powered Visual Renaissance" built using MIT app Inventor whose soul funtion is to send prompt to OpenAi's API to get visuals of the given prompt.This project was created for educational purpose and to learn about how API works.
This repository features a basic implementation of a neural network using TensorFlow and Keras to recognize handwritten digits from the MNIST dataset. The model comprises a flattening layer, two dense layers with ReLU activation, and a final output layer with softmax activation. The script achieves high accuracy on the test set 99.91%.
This repository hosts a robust Image Captioning application, leveraging the powerful Hugging Face API for seamless image-to-text conversion and Gradio to craft an interactive, technically refined interface. By uploading an image, users can harness the model's capabilities to generate precise and contextually rich captions.
This project is concerned with the analysis, design, development, implementation and evaluation of an e-learning management system to provide a user friendly environment for prospective students to acquire knowledge at any educational level and to bridge the gap between teachers and students.
The official Meta Llama 3 GitHub site
This repository was build for my matlab files.
For custom readme for adding it to my profile.
This was our hackathon project where we have to calculate distance using monocular depth estimation technique , we used depth estimation maps for getting the depth of the various features in a 2D image , using MiDas Machine Learning Model and some custom mathematics to calculate rough distances of the features from camera module.
This was our project to provide the online proctored examination facility, with our self build online examination system in php named as "Proctoria" . I along with my co-founder 'Koushik' successfully conducted various tests in our own school.
Segmentation models with pretrained backbones. Keras and TensorFlow Keras.
Code written in Python using the Tkinter library for creating a GUI application. The code is for creating a voice assistant program that can perform various tasks such as searching on Wikipedia, opening web pages, playing music, getting weather information, opening applications, taking screenshots, and more.