Ramiz Hasan Siddiqui

Projects: - Data driven localization: Developed learnable Particle Filter based localization algorithms. Investigating RNNs and VAEs - Corona Warn App: I worked on a proximity detection algorithm based on Bluetooth based distance estimation. Research: - Research Areas: Particle Filters, Generative Models (VAEs), RNNs, Learnable Bayesian Filters - Co-authored a paper (IPIN 2019): “A bidirectional LSTM for estimating dynamic human velocities from a single IMU” - Supervised 1 bachelor and 1 master thesis

Topic: “Indoor Pedestrian SLAM using IMU Step Length Model and UWB Signals“ - Developed an inertial navigation system using XSens IMU for positioning. - Developed a new empirical step length model using IMU sensor data. - Sensor fusion of IMU data and the range measurements from the Ultra wide-band (UWB) transmitters - Developed a Simultaneous Localization And Mapping (SLAM) algorithm to track the pedestrian and also determine the location of UWB sensors. Rao Blackwellized Particle filter was used.

Implementation of Radio Transmitter and Receiver with USRP for Maritime Broadband Communications - Implemented a complete end-to-end transceiver chain using Software Defined radios (USRP B210) to test transmissions and research the optimal communication system. - Coded digital signal processing blocks in C++ in GNURadio for processing blocks for QPSK modulation, OFDM Modulation, channel estimation and equalization. - Took part in measurement campaign at Ammersee to test OFDM symbol synch. over water.

I worked on the “mobisaar Lotse” app which is an app for helping elders in using public transportation. My work on the app included: - I developed navigation feature using OpenStreetMaps and integrated it in the app - Improved and enhanced the user interface (added new entries in scroll menu etc.) - Removed bugs