KSHITIJ SINGH

Student

About

A highly motivated and skilled student with a passion for technology and a strong foundation in computer science. Proven ability to develop innovative solutions and collaborate effectively in team environments. Eager to contribute to challenging projects and make a meaningful impact.

Work

CISMR Lab

|

Research Intern

Summary

A gesture recognition system using Arduino, OpenCV, and Unity 3D, enabling intuitive control of a robotic arm. Leveraged mobile applications, flex sensors, and the MediaPipe library to manipulate robotic arm movements; decreased latency by 15% and increased operational precision by 22% during gesture-controlled tasks.

Highlights

Enabled intuitive control of a robotic arm using Arduino, OpenCV, and Unity 3D.

Leveraged mobile applications, flex sensors, and the MediaPipe library to manipulate robotic arm movements.

Decreased latency by 15% and increased operational precision by 22% during gesture-controlled tasks.

SHUNIYA VIGYAN PRIVATE LIMITED

|

Flutter Developer Intern

Summary

Architected and sustained two high-performance Flutter applications, Bholu App and Library Merchant App, with Bloc state management; achieved 30% faster load times, enhanced user experience. Partnered with a diverse team of designers and backend developers to seamlessly integrate RESTful APIs, significantly enhancing app performance and user experience; achieved a 30% reduction in load times and a 20% increase in user engagement.

Highlights

Architected and sustained two high-performance Flutter applications, Bholu App and Library Merchant App, with Bloc state management.

Achieved 30% faster load times and enhanced user experience.

Partnered with a diverse team of designers and backend developers to seamlessly integrate RESTful APIs.

Significantly enhanced app performance and user experience; achieved a 30% reduction in load times and a 20% increase in user engagement.

Education

National Institute of Technology, Delhi

→
Bachelor of Technology CSE Grade: 7.46

Awards

CAPSULE VISION CHALLENGE

IIITDM

Rank 12th in CAPSULE VISION CHALLENGE organised by IIITDM

Publications

Multi-Class Gastrointestinal Abnormality Detection

Developed a transfer learning model using Swin Transformer to classify gastrointestinal abnormalities from capsule endoscopy images, achieving 93% overall accuracy. Fine-tuned the model on a dataset of 10 classes, demonstrating precision ranging from 0.65 to 0.99 and Fl-scores between 0.50 and 0.98, with notable performance in detecting Ulcer and Worms.

Certificates

Complete A.I. Machine Learning, Data Science Bootcamp

Skills

Languages

C, Python, JavaScript, Dart, SQL.

Frameworks/Libraries

Tensorflow, Scikit-Learn, Pandas, Numpy, OpenCV, Matplotlib, Flutter.

Database

MongoDB, MySql, Firestore.

Developer Tools

Git, GitHub, VS Code.

Interests

Coding Profile

Leetcode - Problem Solved - 750+, Codeforces - Specialist - Rating - 1544(max).

Extra-Curricular

Deputy General Secretary at Robotics Club, Organized and conducted a 5-day robotics workshop and competition at NIT Delhi, featuring hands-on learning, expert guidance, and cutting-edge technologies..

Additional Achievements

NIT DELHI Football Team Captain., Zeal 2023 Best Football Player..

Projects

MNIST Digit Classification

Summary

Developed a CNN model using TensorFlow Keras to classify handwritten digits from the MNIST dataset, achieving 98.5% accuracy. Built an interactive Tkinter-based UI where users can draw digits, and the model provides real-time classification updates.

Image Captioning App

Summary

Developed an image captioning application using Flutter for the front-end, leveraging the Xception model for image feature extraction and LSTM for caption generation. Implemented a Flask-based API for the back-end, ensuring efficient image processing and seamless communication between the app and the server.