About
LakshitUpreti
I'mData Scientist and ML Engineer
About
Namaste !! I'm Lakshit Upreti a Data Science and Machine Learning Enthusiast from Uttrakhand, INDIA. Currently pursuing my degree in Computer Science and Engineering from the prestigious Indian Institute of Information Technology Nagpur. My journey into the realm of data science and machine learning began with a serendipitous discovery during my 5th semester break. Initially dabbling in web development, I soon realized my passion lay elsewhere, ignited by my innate love for mathematics—a trait inherited from my father, a mathematics teacher. In the mean time I also found my love for data, now the path was clear. Delving into the fascinating world of data and mathematics, I made the pivotal decision to pivot towards data science and machine learning. It's where my heart lies, where I find joy in exploring the depths of data and its insights. Beyond models and datasets I love to play badminton and chess.
Services
Here Are Some of My Skills
Services
1 - Data Science and Machine Learning
- Data Analysis
- Data Visualization & PowerBI
- SQL
- Machine Learning & Deep Learning
- Natural Language Processing
3 - Programming Languages
- C++
- Python
- Java
- JavaScript
2 - Front End Development
- HTML / CSS
- JS & Jquery Startup
- WordPress
- Figma
4 - Product Management
- Technical & Non-verbal Communication
- Planning and forecasting
- Team management
- Knowledge of project management methodologies
Works
Projects Done
Portfolio
Assistant
This project has 4 key components: Assistant for real-time responses, Image QA for image-based QA, AskMe for question-answering and MCQ Assistant that generate PDF with MCQs and their answers based on 4 user inputs.
Deep Fake Detection
This project is deep fake detection web application which recognize deep fake audio, video and images very efficiently and effectively. Applied CNN, RNN and anomaly detection techniques, achieving 95% accuracy in detecting deepfakes.
Pokemon Doppelganger
"Pokémon Doppelganger" is an innovative application that employs facial recognition technology and machine learning to find the Pokémon species most closely resembling a user's facial features. By analyzing facial embeddings extracted from user images and comparing them with pre-computed embeddings of Pokémon species, the system identifies the closest match using cosine similarity.
Hand Tracking and Gesture Recognition
This project is on hand tracking and gesture recognition using OpenCV and MediaPipe. The project is designed to detect and track hands in real-time using a webcam, and to recognize hand gestures for various applications, such as controlling the volume of your computer.
Blog
Read Our Case
Our Blog
Understanding ROC Curves: A Beginner’s Journey
Embark on a beginner's journey into ROC curves, exploring their significance and applications in data analysis and classification tasks, gaining foundational insights into this essential tool for assessing classifier performance.
Understanding ROC Curves: A Beginner’s Journey
Embark on a beginner's journey into ROC curves, exploring their significance and applications in data analysis and classification tasks, gaining foundational insights into this essential tool for assessing classifier performance.
Understanding ROC Curves: A Beginner’s Journey
Embark on a beginner's journey into ROC curves, exploring their significance and applications in data analysis and classification tasks, gaining foundational insights into this essential tool for assessing classifier performance.