prakhar-bhartiya Goto Github PK

followers: 5.0 following: 3.0 repos: 29.0 gists: 0.0

Name: Prakhar Bhartiya

Type: User

Bio: Always seeking out new challenges and pushing the boundaries of what's possible.

Location: U.S.

Blog: https://prakhar-bhartiya.github.io/

Hey there!

👨‍💻 About Me :

I am a developer that enjoys brainstorming and attempting to solve problems in many varieties of fields.

Study : Master's in Computer Science @ ASU
Work : Currently looking for full-time roles.
Area : Augmented Reality, Virtual Reality, Mixed/Extended Reality, Game Developement, Machine Learning, Artificial Intelligence, Computer Vision, UI/UX, Data Visualization, Full-stack(intermediate)

🛠️ Languages and Tools :

Programming Language -----> C, C++, C#, Python, HTML, CSS, JavaScript, SQL
Developer Tools -----> Git, VS Code, Unity3D, Blender, SketchUp, AxureRP
SDKs -----> WebXR, Vuforia, ARKit, ARCore, Wikitude, OpenXR, AR.js, Three.js, OpenGL
Libraries -----> TensorFlow, Keras, PyTorch, Scikit-learn, Matplotlib, Pandas, NumPy, OpenCV, OCR, YOLOv3
Hardware -----> Meta Quest2, Meta Quest Pro, Pico Neo 3, Varjo XR-3, Microsoft Hololense, Raspberry Pi, Arduino

✨ Personal Projects :

Virtual Piano : Play piano from any plane surface using Mediapipe Hand tracking system.
Stellar Variability Investigation in VR : NASA - TESS data investigation in VR for Exoplanet discover + Stellar Variability.
Breast Cancer Detection : Brease cancer detection and visualization.
BrainNet : Android application to solve the issue of biometric liveness detection from an EEG dataset.
EMNIST Evaluation : Variants in the Extended MNIST dataset were subjected to in-depth analysis. I was able to achieve comparable accuracy to state-of-the-art models using a simple Conv2D CNN model and precise hyperparameter tuning.
Depression Detection using Sentiment Analysis of social media : Formulated a machine learning model to analyse thousands of Twitter tweets, to predict people’s sentiment by the correlation between emotions and specific linguistic features.
Best Buy : Visualise commercial product data in AR on product by scanning barcode.
Project Trash : Built a CNN model (Resnet101 variant) for garbage classification and volume estimation. Also accurately mapped and visualized data using heat maps. Citizens can provide an image dataset, later used by local municipalities.
ShopAR : Try apparels on your 3D model(using single picture) before buying online in Augmented Reality.
AR Assist : Augmented reality training program for manufacturing products.
ProHelm : Smart biker's helmet with XR (Extended Reality) System for visual assistance in navigation(HUD). Extended augmented reality-based display effects were achieved by reflecting a mini projector onto the translucent helmet visor at a certain angle. Results provide a hands-free and distraction less biking experience.
MUnix : Mobile + UNIX = MUNIX is a featured phone interface build on a UNIX system using command and Perl scripting for implementing various features of the phone.

✨ Other Projects :

Bi Direction Search - PacMan : Bi-Direction Search, DFS, BFS, UCS and A-Star implementation for Pacman game.
Image Retrival System : Developed an image retrieval system. Color Moments, LBP, and HOG are examples of implemented feature extraction methods. Then, using techniques like PCA, LDA, SVD, and K-means, reduce the dimensionality to the top K features. Decision Tree, SVM, and Personalised PageRank classifiers were used to classify the data.

Prakhar Bhartiya's Projects

ar-assist

AR ASSIST is an augmented reality-based solution in industries for providing an alternative more human-friendly way of interactions & learning experience of various products and components. It also facilitates new trainees in the early stages of training. Model Recognition, Plane Detection, 3D Point Cloud opt for better customer experience.

blender

Contains blender models

brainnet

An android application to detect the liveness of EEG signals. This application takes in user input and selects data from a provided data set to predict its liveness. The app communicates to a server via the Internet to retrieve the data and machine learning models.

breast_cancer_detection

Breast Cancer Detection Using CNN

dlaicourse

Notebooks for learning deep learning

easy-grid

easy-grid is a React component factory that provides a declarative layout mechanism for utilizing CSS grid layouts.

emnist_evaluation_cse598_project

Evalution and analysis on EMNIST dataset with CNN. Project was created for CSE 598 : Intro Deep Learning Fall'21 @ASU

imitan-navkiran_sathi

A ChatGPT(GPT-3) application for helping Navkiran students.

instant-ngp-test

Instant neural graphics primitives: lightning fast NeRF and more

ml-agents

Unity Machine Learning Agents Toolkit

opencv-learn

Repo for OpenCV Learned codes and experimental results.

phase2

pixel_nasa-stellar-variability-investigation-in-vr

The goal of this project is to demonstrate how anyone, even those without any prior astronomical knowledge, can learn about the study of star variability using the well-known method of "light curves" inspection in a virtual reality environment.

pixi_best_buy

Augmented reality visualization tool for commercial products. Integrated features like the product's 3D model visualizer with information and relevant health tips and in-built indoor navigation for better alternative recommended products. This idea is well suitable for supermarkets providing an enhanced shopping experience (Database: GS1 Database)