Bernd Pfrommer

Bernd Pfrommer

Visiting Scholar

University of Pennsylvania


I am a Visiting Scholar at the University of Pennsylvania, working in the group of Professor Kostas Daniilidis. My research interests include computer vision, autonomous robots, and machine learning.


  • Computer Vision
  • Autonomous Systems
  • Machine Learning


  • PhD in Physics, 1998

    University of California at Berkeley

  • MSc in Computer Science, 1998

    University of California at Berkeley

  • Diploma in Physics, 1993

    Universität Stuttgart



Postdoctoral Researcher and Visiting Scholar

University of Pennsyvlania Engineering / GRASP Laboratory

Dec 2015 – Present Philadelphia
Working as a researcher and developer on several projects related to:

  • Computer vision
  • Machine learning
  • Visual-inertial odometry
  • Camera calibration

Founder and Managing Partner

Black Forest Trading LLC

Jul 2002 – Jan 2012 Connecticut
Responsibilities include:

  • Researching high frequency financial trading strategies
  • Managing day-to-day trading operations
  • Writing software for fully automated electronic trading of equities and futures
  • Running a company (accounting, taxes, contract negotiations and other fun stuff)

Electronic Market Making Researcher


May 2000 – Jun 2002 Connecticut
Developed short term trading strategies and lead team of 2-4 researchers.

Senior Developer, Program Trading

Bank of America

Sep 1999 – May 2000 San Francisco
Wrote C++/Java software for a BofA’s Program Trading desk.

Member of Technical Staff III and IV

Sun Microsystems

Feb 1998 – Sep 1999 Menlo Park
Wrote C/C++ software for MPI high performance message passing library and kernel components.




Flexible SLAM with tags

UPenn Smart Aviary

Tracking birds in 3D


TagSLAM: Robust SLAM with Fiducial Markers

TagSLAM provides a convenient, flexible, and robust way of performing Simultaneous Localization and Mapping (SLAM) with AprilTag fiducial markers.

Temperature-Dependent Hydrogen-Bond Geometry in Liquid Water

We have determined the hydrogen-bond geometry in liquid water from 0 to 80C by combining measurements of the proton magnetic shielding tensor with ab initio density functional calculations.

Si-O-Si bond-angle distribution in vitreous silica from first-principles 29 Si NMR analysis

The correlation between 29 Si chemical shifts and Si-O-Si bond angles in SiO2 is determined within density-functional theory for the full range of angles present in vitreous silica

NMR Chemical Shifts of Ice and Liquid Water:  The Effects of Condensation

We report the results of ab initio density functional theory calculations of the NMR chemical shift of liquid water and hexagonal ice.

Signature of surface states on NMR chemical shifts: A theoretical prediction

We show with an ab initio calculation that electronic surface states have a strong effect on NMR chemical-shift spectra.