01.
I have always been curious to learn how things work, the engineering in small things is very intriguing to me. I like being involved in making new things, be it my first transistor based circuit in 5th standard or the Machine Learning based projects I have been doing since last two years. I have keen interest in robotics and the love started back in 11th standard in a workshop during my high school and there has been no looking back since then. I am working with constant perseverance with one sole purpose of learning robotics in my mind. In the first year of my engineering I studied about Arduino, Raspberry pi, AVR, ARM and made various projects using them.
Then, I went on to explore the applications of deep learning and computer vision in robotics. The field is so immersive and has a lot to explore, I'm working on improving my concepts and projects in deep learning since then. I have worked on various topics of deep learning including computer vision, nlp, time series forecasting etc. but, most of my work is in the field of computer vision based on deep learning as, it interests me more and involves a lot of robotics. In computer vision I have worked on various projects some of them can be found below in my work section. I am also working on a research paper with Associate Professor in my college. Recently, I have been studying deep reinforcement learning algorithms and have implemented DQN based pong AI.
In future, I look forward to work and learn more in the field. I want to contribute to the open source deep learning community, so that it reaches to more and more engineers. The main goal is to implement this knowledge for making life simpler and better.
Education
Bachelor of Engineering
08.2016 - 06.2020
Thapar Institute of Engineering and Technology
Major - Electronics and Communication Engineering
Current CGPA- 8.07/10
Education
High School
2015 - 2016
DAV Public School, Panipat
Percentage: 90.2
Education
Secondary School
2013 - 2014
Birla Public School, Pilani
CGPA: 9.4/10
Experience
General Secretary
IETE
General Secretary Managing day to day work of thapar chapter of nationwide technical society of electronics and communication. I also mentor 24 students under me. IETE, Patiala
Technical Head
IETE
I arranged workshops on several stuffs such as arduino and its peripherals. I also volunteered to mentor 15 students through their freshman year for technical stuff
Technical Head
Elementos
Technical head of a week long tech fest of robotic, electronics. I also managed a hackathon competition in the same fest.
Hobbies
Music
Blogging
Badminton
Gaming
02.
Skills
Languages
- Python
- C++
- Matlab
Deep Learning
- Pytorch
- Tensorflow
- Keras
- DQN
- DDQN
- A3C
Deep Reinforcement Learning
Robotics and Electronics
- Arduino and peripherals
- Raspberry pi
- Vrep(simulation tool)
- Embedded system
Other Skills
- Image Processing
- Git
- Latex
- Scikit-learn
- Librosa
- Opencv
03.
My Work
Attention Detection
Research project in which we are detecting the engagement level of students during MOOC using deep learning. (Ongoing)
Crime Detection
Detecting crimes from CCTV footage using UCF crime datasetusing MIL ranking algorithm and slow-fast networks for feature extraction. This project was done as a freelance work for a mexican company
Artificial eyes
Device for blinds that uses CNN and LSTM to generate caption and converts the caption of the image to speech and get output on raspberry pi. Similar thing can be implemented using mobile phones to make it more accessible.
Self Driving Car
Implementing using simulation tools, deep learning, and other computer vision techniques.(Ongoing Lane finding, traffic sign classifier and steering angle prediction implemented)
GAN Applications
I am implementing various applications of GANs. I have implemented DCGAN and semantic segmantation using pix2pix GAN.
Try it
HealthCad
Implemented some of the latest deep learning algorithms to help doctors diagnose various diseases at an early stage and make it accessible by deploying on cloud platform.
Try it
Music genre classification
Genre classification of music using FMA small dataset. I first took the fft of song sequence and created spectogram. I applied cnn(Densenet) on the spectograms created.
Try it
Neural Style transfer
Artistic style transfer based on the research paper by Gatys. Made it using VGG16 architecture and pytorch. I tried few other combinations other than those given in the paper but it worked best in case of those suggested in paper.
Age and gender detection
Age and gender detection of a person from the image of face using UTK face dataset. I applied transfer learning using resnet18 architecture in pytorch
Music Lyrics Generation
Eminem like lyrics generation using LSTM networks. Lyrics were scrapped using pylyrics3 library and keras, nltk was used to preprocess data and create lstm.
Home automation using IOT
Home automation using node-mcu and cayenne cloud platform. Applied it in my room and used it for a week using an app of cayenne.
Line follower using PID
Line follower based on PID algorithm to improve accuracy. Participated in saturnalia with this bot and reached finals.
Seed sowing bot
Seed sowing bot model using arduino and ir sensors. Used servo motors for making seed sowing mechanism and used ir sensors to detect pits.