PRAYUSHI MATHUR

Biography

Currently pursuing Computer Engineer from Nirma University, Ahmedabad, India. Worked across diverse domains like Computer Vision, Image Processing, Medical Imaging, Satellite Imagery, Generative and Adversial Networks in the field of Artificial Intelligence and Machine Learning from the past 1.5 years. I have been associated with various clubs and societies for organizing various events. Apart from academic activities, I am a trained classical singer and Ex-Joint Secretary of ISTE Students’ Chapter.

Education

• Bachelor’s - Nirma University, Ahmedabad
• High School - Shiv Jyoti Convent School, Kota
• Primary and Secondary School - Sophia Girls’ Senior Secondary School, Kota

Skills

• Programming Languages: Python, C
• Tools: Anaconda, Matlab, Jupyter Notebook, Eclipse, Xampp, Spyder
• Frameworks: Tensorflow, Keras, Scikit-Learn
• System: Windos(98/vista/7/8/8.1/10), Linux, Ubuntu

Research

Currently, I am involved in a research project of Multiple Organ Segmentation and NLP using Deep Learning.

3D Glioma Brain Tumor Segmentation of MR Images

• Segmentation of glioma tumor from MRI considering 3D volume.
• Our model trained on BraTS 2018-19 dataset.

Denoising Autoencoder using GAN

• Combined the two prominent models of deep learning i.e. Autoencoder and GAN.
• Denoising the MNIST dataset and 2-D Brain image dataset.
• Our model used Stacked Sparse Denoising Autoencoder and reached to a good PSNR value.

Object Identification

• Implemented YOLO v2/v3 on custom objects which converts high resolution images to low resolution images and try to detect the objects.
• Worked on Retinanet (RESNET50) which maintains the quality of image.
• Currently working on Satellite Images.

Multi-organ Segmentation

• Performed multi-organ segmentation on 8 organs of the human body.
• The organs included the pancreas, three organs fronm gastrointestinal tract (esophagus, stomach and duodenum), liver, gallbladder, spleen and left kidney.

Type-2 Diabetes

• Worked in Data Mining and Machine Learning on real data of around 12k patients containing 373+ attributes in collaboration with Diacare Diabetes Center.
• It included the triopathy of diabetes i.e. Nephropathy, Neuropathy and Retinopathy.
• The model act as an assistant of the doctor which helps in diagnosis of Nephropathy mainly.

Other Projects

Music Creator

• Built an Android application for creating and playing musical instruments like piano, drum, etc.
• It also allowed recording various tunes for playback later using Android Studio 2.3.3.

Certifications and Courses

• Deep Learning Specialization, Prof. Andrew Ng.
• Machine Learning by Stanford University, Prof. Andrew Ng.
• Image Processing and Pattern Recognition, NPTEL
• Deep Learning, NPTEL
• Python, IIT Bombay
• Android App Development, IIM Bangalore
• Cloud Computing, IIT Delhi

Hobbies & Interests

• Singing and Listening Songs.
• Reading Novels.
• Surfing the internet.
• Watching TV series.