Muazzam Ali

At Alten I was responsible for solving problems. Some major problems I solved were. Fixing containers inside large trucks optimally. I solved the problem using linear programming. Writing KPIs in C++ for large CV datasets, I used xtensor to implement the solution. For tracking people in live videos I implemented a statistical feature matching histogram comparison algorithm.

• Solve the problem of less labeled datasets by training deep networks in a semi-supervised manner using Tensorflow. • Predicting end-to-end optical flow values using depth information of the video sequences. • Introduced Unsupervised losses to train the network in a semi-supervised manner. • Introduced GAN architecture in FlowNet by treating flowNet as a generator. • Comparisons between accuracy of multiple supervised and unsupervised losses working together.

• Run a trained AutoEncoder from Tensorflow python inside Tensorflow C++. • Python Code optimization in core products implemented with Pandas Dataframe with unit tests. • Extract useful Features from large OpenData Maps dataset using PySpark for Volkswagen. • Implementation of build pipelines using Jenkins. • Implementation of backend architecture using django Rest Api for Consus-Global. • Implementation of frontends using Angular 5 with NgRx architecture.

• Conducting benchmarking experiments for FPS on different hybrid apps frameworks. • Planning and implementation of architecture for building and deploying performant hybrid apps. • Using React-Native with Protocol Buffers to implement libraries for smooth communication with CAN devices. • Building native kernel apps for multiple platforms that acts as Play store for customer based apps.

The aim of the project was to calculate useful metrics from large RDF graphs. My work included calculation of node importance using PageRank and node Centralities. For centrality measurement, I implemented algorithms for Single Source Shortest Path (SSSP) on undirected unweighted graphs . The project was implemented using Apache Spark framework in Java.

Working as a Software engineer responsible for building mobile apps for different customer based products.

Responsible for implementing features for their customer based software products. Working on different frontend and backend web technologies and frameworks like Ruby on Rails, EmberJS, Node.js, Android

The project was a lab course during the University. The aim of the project was to identify free parking spots autonomously which we achieved by mapping 3D data using Point cloud library(PCL) to 2D images. After which we applied logistic regression on parking spots for accurate detection. The development for the project was done on C++.

In the Computer vision lab, I was working as a work student where I was responsible for developing a python application in blender that will convert and take snapshots of different 3D models with different camera and lighting properties. The application was further supposed to be used for a Master thesis by another student.

Machine learning, Robotics, AI, Computer Vision, Big Data