An Open Source contributor with Master’s degree in Data Science and 3 plus years of hands-on experience in the field of Artificial Intelligence. Where the primary focus is on Machine Learning, Computer Vision, and Natural Language Processing applications.
For Computer Vision I have mostly worked on Object Tracking, Real Time Segmentation, and some of its advanced techniques such as Deep Image Matting and Salient Object Detection.
For Natural Language Processing I have good understanding of Transformers, and how we can efficiently deploy them. Currently, I am further exploring NLP and working on NLU and NLG tasks.
With future interest in CV, NLP and RL, I am looking for a remote role with a diverse background, where I can learn and contribute to the organization's growth.
Working on an NLP project called News Revealer. Which is a platform for news generation in which users can summarise, paraphrase, and generate articles. Currently deploying transformers-based models using Flask, Docker, NGINX, and NVIDIA Triton Server.
Skills: Natural Language Processing (NLP) · Docker · DevOps · Python (Programming Language) · Deep Learning
Worked on two Computer Vision and one Data Science project.
Nails Try On: A segmentation based task in which using a state of the art real time custom segmentation model I was able to achieve 13% improvement in accuracy and around 20% improvement in terms of speed.
Beauty: An image to image translation task in which by changing input flow I was able to relatively reduce 30% overfitting in existing cGAN and Pix2Pix GAN model.
Luccro: A decision support engine for financial trading(Forex) that uses AI to identify low-risk, high-reward opportunities for traders. Generated trading indicators, and worked on model training and deployment of Technical and Strength time series data.
Skills: Time Series Analysis · Image Segmentation · Generative Adversarial Networks (GANs) · Research and Development (R&D) · Computer Vision · TensorFlow · Python (Programming Language) · Deep Learning · Machine Learning
Developed three android applications in Java.
Parliamentary Proceedings: Application designed to observe the proceedings of the Senate, National Assembly, and 13 selected Standing Committees under project Parliament Watch and Reforms Project (PWRP).
FAFEN Observation: A Survey Application that collected the data of the 2018 General Elections held in Pakistan under Citizens' Voice Project (CVP).
Pooling Station Survey: An android application was developed for Election Commission of Pakistan to collect data on potential pooling stations.