Florian T. Pokorny

Assistant Professor, Machine Learning, RPL, EECS, KTH

Welcome

My current research focuses on two main directions: 1) explainable machine learning algorithms with a geometric or topological flavour and 2) machine learning methods that are tailored for robotic manipulation or motion planning and which incorporate available domain knowledge and information about physics and configuration space geometry in order to be data efficient.

Positions & Professional Interests

  • Assistant Professor, Machine Learning, RPL, EECS, KTH
  • Responsible for WASP AI Graduate School at KTH

Short Biography

I am currently an Assistant Professor in Machine Learning within the division of Robotics, Perception and Learning at KTH Royal Institute of Technology. In 2015-2016, I was a postdoctoral visiting scholar at the AMPlab and the Berkeley Automation Science Lab at the University of California, Berkeley where I worked with Ken Goldberg and his group. Previously, I was fortunate to be a researcher and postdoc working with Prof. Danica Kragic and her group. Before then, I was a pure mathematician and completed my PhD entitled "The Bergman Kernel on Toric Kähler Manifolds" under supervision of Prof. Michael Singer at the University of Edinburgh. Before that, I received my BSc mathematics from the University of Edinburgh and a Master of Advanced Study in Mathematics (Part III) from the University of Cambridge and also spent a year as an exchange student at the National University of Singapore.

Contact

Room 713, Teknikringen 14 RPL, EECS, KTH, 114 28 Stockholm, Sweden, fpokorny (at) kth.se

News

Co-organizing ICRA 2019 Workshop on Topological Methods
New course on Topological Data Analysis for WASP students in 2019

Research Group

Current Students - Main Supervisor

Vladislav Polianskii

PhD Student

Geometric and Topological Machine Learning Methods

website

Robert Gieselmann

PhD Student

Machine Learning for Deformable Object Manipulation

 

Fereidoon Kamali

Project Student, DD2411

Machine Learning for Motion Planning

 

Anton Lundberg

Project Student, DD2411

Machine Learning for Motion Planning

 

Current Students - Co-Supervisor

Anastasiia Varava

PhD Student, main supervisor: Danica Kragic

Caging and Motion Planning Algorithms

website

Joao Carvalho

PhD Student, main supervisor: Danica Kragic

Topological Methods for Robotics

website

Michael Welle

PhD Student, main supervisor: Danica Kragic

Robotic Manipulation

website

Rita Laezza

PhD Student (Chalmers), main supervisor: Yiannis Karayiannidis

Deformable Object Manipulation

website

Past Supervision and Co-supervision

Marcel Binz

Master Student, 2017

Thesis: "Learning Goal Directed Behaviour"

Magnus Raunio

Master Student with Mikael Vejdemo Johansson, 2014

Thesis: "Periodic Motion Extraction Using Harmonic Regression and Structural Parameterization"

Kaiyu Hang

cosupervised PhD student, main supervisor: Danica Kragic

Thesis: "Dexterous Grasping – Representation and Optimization"

next: Assist. Prof. HKUST, postdoc Yale.

website

Püren Güler

cosupervised PhD student for part of PhD, main supervisor: Danica Kragic

next: postdoc Örebro University

website

Andrea Baisero

Swedish AI Society's best master thesis award

Master Student joint with Carl Henrik Ek and Danica Kragic

Next: PhD student University of Stuttgart and Northeastern University

website, thesis

Patrick Truong

Master Student

Next: PhD student at KTH

website

Vladislav Polianskii

Master Student

Next: PhD student under my supervision

website

Publications

See also google scholar

2019

Voronoi Classification: A High-Dimensional Geometric Approach via Weighted Monte Carlo Integration over Voronoi Boundaries

Vladislav Polianskii, Florian T. Pokorny

International Conference on Machine Learning (ICML), 2019

pdf, appendix

2018

Free Space of Rigid Objects: Caging, Path Non-Existence, and Narrow Passage Detection

Anastasiia Varava, J. Frederico Carvalho, Florian T. Pokorny, Danica Kragic

Workshop on Algorithmic Foundations of Robotics (WAFR) 2018

pdf, appendix

Synthesis of Energy-Bounded Planar Caging Grasps using Persistent Homology

Jeffrey Mahler, Florian T. Pokorny, Sherdil Niyaz, Ken Goldberg

IEEE Transactions on Automation Science. Volume 15, Issue 3. 2018

pdf, website

Path Clustering with Homology Area

J. Frederico Carvalho, Mikael Vejdemo-Johansson, Danica Kragic, Florian T. Pokorny

IEEE ICRA 2018

pdf

SWIRL: A sequential windowed inverse reinforcement learning algorithm for robot tasks with delayed rewards

Sanjay Krishnan, Animesh Garg, Richard Liaw, Lauren Miller, Florian T. Pokorny, Ken Goldberg

The International Journal of Robotics Research (IJRR). Volume 37. Issue 9. July, 2018

pdf

An algorithm for calculating top-dimensional bounding chains

J. Frederico Carvalho, Mikael Vejdemo-Johansson, Danica Kragic, Florian T. Pokorny

PeerJ Computer Science

pdf

2017

Caging and Path Non-Existence: a Deterministic Sampling-Based Verification Algorithm

Anastasiia Varava, J. Frederico Carvalho, Florian T. Pokorny, Danica Kragic

International Symposium on Robotics Research, 2017

pdf

Caging Grasp Synthesis and Verification for Partially Deformable 3D Objects with Narrow Parts

Anastasiia Varava, Danica Kragic, Florian T. Pokorny

IEEE IROS (Abstract Only), 2017

Multi-scale Activity Estimation with Spatial Abstractions

Majd Hawasly, Florian T. Pokorny, Subramanian Ramamoorthy

Geometric Science of Information, 2017

pdf

Herding by Caging: a Topological Approach towards Guiding Moving Agents via Mobile Robots

Anastasiia Varava, Kaiyu Hang, Danica Kragic, Florian T. Pokorny

Proceedings of Robotics: Science and Systems, 2017

pdf

CapriDB - Capture, Print, Innovate: A Low-Cost Pipeline and Database for Reproducible Manipulation Research

Florian T. Pokorny, Yasemin Bekiroglu, Karl Pauwels, Judith Bütepage, Clara Scherer, Danica Kragic

Data in Brief, Volume 11, April 2017, Pages 491-498

pdf

2016

Large-Scale Supervised Learning of the Grasp Robustness of Surface Patch Pairs

Daniel Seita, Florian T. Pokorny, Jeffrey Mahler, Danica Kragic, Michael Franklin, John Canny, Ken Goldberg

IEEE SIMPAR, 2016

pdf

Synthesis of Energy-Bounded Planar Caging Grasps using Persistent Homology

Jeffrey Mahler*, Florian T. Pokorny*, Sherdil Niyaz, Ken Goldberg

WAFR, 2016

pdf

SWIRL: A Sequential Windowed Inverse Reinforcement Learning Algorithm for Robot Tasks With Delayed Rewards

Sanjay Krishnan, Animesh Garg, Richard Liaw, Brijen Thananjeyan, Lauren Miller, Florian T. Pokorny, Ken Goldberg

WAFR, 2016

pdf

Caging Grasps of Rigid and Partially Deformable 3D Objects with Double Fork and Neck Features

Anastasiia Varava, Danica Kragic, Florian T. Pokorny

IEEE Transactions on Robotics, 2016

pdf

Hierarchical Fingertip Space: A Unified Framework for Grasp Planning and In-Hand Grasp Adaptation

Kaiyu Hang, Miao Li, Johannes A. Stork, Yasemin Bekiroglu, Florian T. Pokorny, Aude Billard, Danica Kragic

IEEE Transactions on Robotics, 2016

pdf

Privacy-Preserving Cloud-Based Grasp Planning

Jeffrey Mahler, Brian Hou, Sherdil Niyaz, Florian T Pokorny, Ramu Chandra, Ken Goldberg

IEEE CASE, 2016, Finalist, Best Student Paper Award, CASE 2016

pdf

Robot Grasping in Clutter: Using a Hierarchy of Supervisors for Learning from Demonstrations

Michael Laskey, Jonathan Lee, Caleb Chuck, David Gealy, Wesley Hsieh, Florian T Pokorny, Anca D Dragan, Ken Goldberg

IEEE CASE, 2016

pdf

High-Dimensional Winding-Augmented Motion Planning with 2D Topological Task Projections and Persistent Homology

Florian T Pokorny, Danica Kragic, Lydia E Kavraki, Ken Goldberg

IEEE ICRA, 2016

pdf

Topological Trajectory Clustering with Relative Persistent Homology

Florian T Pokorny, Ken Goldberg, Danica Kragic

IEEE ICRA, 2016

pdf

Dex-Net 1.0: A Cloud-Based Network of 3D Objects and a Multi-Armed Bandit Model with Correlated Rewards to Accelerate Robust Grasp Planning

Jeffrey Mahler, Florian T. Pokorny, Brian Hou, Melrose Roderick, Michael Laskey, Mathieu Aubry, Kai Kohlhoff, Torsten Kroeger, James Kuffner, Ken Goldberg

IEEE ICRA, 2016, Finalist, Best Manipulation Paper Award, ICRA 2016

pdf

CapriDB - Capture, Print, Innovate: A Low-Cost Pipeline and Database for Reproducible Manipulation Research

Florian T. Pokorny, Yasemin Bekiroglu, Karl Pauwels, Judith Bütepage, Clara Scherer, Danica Kragic

Workshop on Grasping and Manipulation Datasets, IEEE ICRA, 2016

See the 2017 Data in Brief Journal publication for details

Energy-Bounded Caging: Formal Definition and 2D Lower Bound Algorithm Based on Weighted Alpha Shapes

Jeffrey Mahler, Florian T. Pokorny, A. Frank van der Stappen, Ken Goldberg

IEEE RA-Letters, 2016

pdf

SHIV: Reducing Supervisor Burden using Support Vectors to Efficiently Learn Robot Control Policies from Demonstrations

Michael Laskey, Sam Stazak, Wesley Yu-Shu Hsieh, Florian T. Pokorny, Anca D. Dragan, Ken Goldberg

IEEE ICRA, 2016, Finalist, Best HRI Paper Award, ICRA 2016

pdf

TSC-DL: Unsupervised Segmentation of Surgical Demonstrations with Deep Learning

Adithyavairavan Murali, Animesh Garg, Sanjay Krishnan, Florian T. Pokorny, Pieter Abbeel, Trevor Darrell, Ken Goldberg

IEEE ICRA, 2016

pdf

Topological Trajectory Classification with Filtrations of Simplicial Complexes and Persistent Homology

Florian T. Pokorny, Majd Hawasly, Subramanian Ramamoorthy

International Journal of Robotics Research (IJRR), 2016

pdf

2015

On Visual Feature Representations for Transition State Learning in Robotic Task Demonstrations

Animesh Garg*, Sanjay Krishnan*, Adithyavairavan Murali, Florian T. Pokorny, Pieter Abbeel, Trevor Darrell, Ken Goldberg

Advances in Neural Information Processing Systems 28, Workshop on Feature Extraction: Modern Questions and Challenges, 2015

pdf

Data-Driven Topological Motion Planning with Persistent Cohomology

Florian T. Pokorny, Danica Kragic

Proceedings of Robotics: Science and Systems, 2015

pdf

Multi-Armed Bandit Models for 2D Grasp Planning with Uncertainty

Michael Laskey, Jeff Mahler, Zoe McCarthy, Florian T. Pokorny, Sachin Patil, Jur van den Berg, Danica Kragic, Pieter Abbeel, Ken Goldberg

Proc. of the IEEE International Conference on Automation Science and Engineering (CASE), 2015

pdf

On a Family of Decomposable Kernels on Sequences

Andrea Baisero, Florian T. Pokorny, Carl Henrik Ek

arXiv, 2015

pdf

The Path Kernel: A Novel Kernel for Sequential Data

Andrea Baisero, Florian T. Pokorny, Danica Kragic, Carl Henrik Ek

Pattern Recognition Applications and Methods, 2015, pp. 71-84

pdf

Cohomological learning of periodic motion

Mikael Vejdemo-Johansson, Florian T. Pokorny, Primoz Skraba, Danica Kragic

Applicable Algebra in Engineering, Communication and Computing, 2015, pp. 1-22

pdf, video

2014

Grasp Moduli Spaces, Gaussian Processes, and Multimodal Sensor Data

Florian T. Pokorny, Yasemin Bekiroglu, Johannes Exner, Mårten Björkman, Danica Kragic

Workshop on Information-based Grasp and Manipulation Planning, Robotics: Science and Systems, 2014

pdf, video

Multiscale Topological Trajectory Classification with Persistent Homology

Florian T. Pokorny, Majd Hawasly, Subramanian Ramamoorthy

Proceedings of Robotics: Science and Systems, 2014

Selected for a 20 min oral presentation at RSS (1 of 6 such presentations)

pdf

Grasp Moduli Spaces and Spherical Harmonic

Florian T. Pokorny, Yasemin Bekiroglu, Danica Kragic

Proc. of the IEEE International Conference on Robotics and Automation (ICRA), 2014

pdf, video, code

Combinatorial Optimization for Hierarchical Contact-level Grasping

Kaiyu Hang, Johannes A. Stork, Florian T. Pokorny, Danica Kragic

Proc. of the IEEE International Conference on Robotics and Automation (ICRA), 2014

Finalist, Best Student Paper Award, ICRA 2014

pdf

2013

Supervised Hierarchical Dirichlet Processes with Variational Inference

Cheng Zhang, Carl Henrik Ek, Xavi Gratal, Florian T. Pokorny, Hedvig Kjellström

IEEE ICCV Workshop on Inference for Probabilistic Graphical Models, 2013

pdf

Toric partial density functions and stability of toric varieties

Florian T. Pokorny, Michael Singer

Mathematische Annalen, 2013

pdf, video

A Topology-based Object Representation for Clasping, Latching and Hooking

Johannes A. Stork, Florian T. Pokorny, Danica Kragic

IEEE-RAS International Conference on Humanoid Robots (HUMANOIDS), 2013

Finalist, Best Conference Paper Award, HUMANOIDS 2013

pdf

Classical Grasp Quality Evaluation: New Theory and Algorithms

Florian T. Pokorny, Danica Kragic

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2013

pdf, code

Caging Complex Objects with Geodesic Balls

Dmitry Zarubin, Florian T. Pokorny, Marc Toussaint, Danica Kragic

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2013

pdf

Integrated Motion and Clasp Planning with Virtual Linking

Johannes A. Stork, Florian T. Pokorny, Danica Kragic

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2013

pdf

Friction Coefficients and Grasp Synthesis

Kaiyu Hang, Florian T. Pokorny, Danica Kragic

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2013

pdf

Grasp Moduli Spaces

Florian T. Pokorny, Kaiyu Hang, Danica Kragic

Proceedings of Robotics: Science and Systems, 2013

pdf, video

Towards Postural Synergies for Caging Grasps

Johannes A. Stork, Florian T. Pokorny, Danica Kragic

Hand Synergies - how to tame the complexity of grapsing, Workshop, IEEE International Conference on Robotics and Automation (ICRA), 2013

pdf

Topological Synergies for Grasp Transfer

Dmitry Zarubin, Florian T. Pokorny, Dan Song, Marc Toussaint, Danica Kragic

Hand Synergies - how to tame the complexity of grapsing, Workshop, IEEE International Conference on Robotics and Automation (ICRA), 2013

pdf

Grasping Objects with Holes: A Topological Approach

Florian T. Pokorny, Johannes A. Stork, Danica Kragic

Proc. of the IEEE International Conference on Robotics and Automation (ICRA), 2013

pdf

The Path Kernel

Andrea Baisero, Florian T. Pokorny, Danica Kragic, Carl Henrik Ek

ICPRAM: International Conference on Pattern Recognition Applications and Methods, 2013

pdf

2012

Topological Constraints and Kernel-Based Density Estimation

Florian T. Pokorny, Carl Henrik Ek, Hedvig Kjellström, Danica Kragic

Advances in Neural Information Processing Systems 25, Workshop on Algebraic Topology and Machine Learning, 2012

pdf

Persistent Homology for Learning Densities with Bounded Support

Florian T. Pokorny, Carl Henrik Ek, Hedvig Kjellström, Danica Kragic

Advances in Neural Information Processing Systems 25, 2012, pp. 1826--1834

pdf

2011

Toric partial density functions and stability of toric varieties

Florian T. Pokorny, Michael Singer

ArXiv Preprint

pdf, video

The Bergman Kernel on Toric Kähler Manifolds

Florian T. Pokorny

PhD thesis, The University of Edinburgh, 2011

pdf, bibtex, abstract

Thesis Projects

I currently offer master thesis projects in these topics

If you are interested in any of the topics below or would like to suggest your own topic, drop me an email with your CV and transcripts attached as well as a reason for why you are interested in this topic. Please note that I only have capacity to consider students that are already registered at KTH for thesis project work.

Software and Data

The projects HarmonicRegression and FastGrasp are available

HarmonicRegression (C++/Python)

This C++ library with python bindings implements spherical harmonics and an associated least-squares regression to recover a smooth surface description from point-cloud data. These methods are described in our paper "Grasp Moduli Spaces and Spherical Harmonics", F. T. Pokorny, Y. Bekiroglu, D. Kragic, IEEE ICRA, 2014 This software is freely available for academic/research purposes, just drop me an email to get the code: fpokorny (at) kth.se

Introduction

Consider the space $L^2(\mathbb{S}^2)$ of integrable functions on the sphere $\mathbb{S}^2\subset \mathbb{R}^3$. We choose a coordinate chart $p(\theta, \phi) = (\sin \theta \cos \phi, \sin \theta \sin \phi, \cos\theta)$ on the sphere, with $(\theta, \phi) \in [0, \pi]\times [0, 2\pi)$. For integers $l, m$, the real valued spherical harmonic function of degree $l$ and order $m$, $|m|\le l$, is defined by
\[ Y_{l,m}(\theta, \phi) = \begin{cases} c_{l,m}P_l^{|m|}(\cos \theta)\sin (|m|\phi) & -l\le m \le -1\\ \frac{c_{l,m}}{\sqrt{2}}P_l^0(\cos \theta) & m = 0\\ c_{l,m}P_l^{m}(\cos \theta)\cos(m\phi) & 1\le m \le l, \end{cases} \]
where $c_{l,m} = \sqrt{\frac{2l+1}{2\pi}\frac{(l-|m|)!}{(l+|m|)!}}$ and $P_l^m$ denotes the associated Legendre polynomial of order $m$ and degree $l$. The functions $Y_{l,m}$ arise as eigen-functions of the Laplacian $\Delta=\frac{1}{\sin\theta}\frac{\partial}{\partial\theta}(\sin\theta \frac{\partial}{\partial\theta})+\frac{1}{\sin^2\theta}\frac{\partial^2}{\partial^2\phi}$ on $\mathbb{S}^2$ For fixed $l$, there are $2l+1$ eigenfunctions $\{Y_{l,m}: m\in\{-l, \ldots, l\}\}$ satisfying \[ \Delta Y_{l,m} = \lambda_l Y_{l,m}, \] where $\lambda_l = l(l+1)$ denotes the corresponding eigen-value. These eigen-functions, for all $l, m$, form an infinite orthonormal basis for $L^2(\mathbb{S}^2)$ with respect to the standard inner product $\langle f, g\rangle = \int_{\mathbb{S}^2} fg \,dVol$, for $f, g\in L^2(\mathbb{S}^2)$ and where $dVol = \sin \theta d \theta d \phi$.

A visualization of the functions $Y_{l, m}$, where the degree changes from 0 to 4 along the horizontal axis and the order $m$ changes from $-l$ to $l$ along the vertical axis.

This library can be used to compute the functions $Y_{l, m}$ and their derivatives. Additionally, a least squares regression of point-cloud data which can be mapped onto a sphere can be performed easily with the provided code. The figure below illustrates an example reconstruction of a smooth surface from point-cloud data where all $Y_{l, m}$ up to degree $L$ are used.

Point cloud P with 25000 points and reconstructed surfaces, where all eigen-spaces up to degree $L$ are used.

FastGrasp (C++/Python)

This C++ library with python bindings implements fast grasp quality evaluation methods outlined in our paper "Classical Grasp Quality Evaluation: New Theory and Algorithms", F. T. Pokorny, D. Kragic, IEEE/RSJ IROS, 2013. This software is freely available for academic/research purposes, just drop me an email to get the code: fpokorny (at) kth.se

Introduction

This library is concerned with the evaluation of a classical $L^1$ grasp quality score $Q(g)$ for a grasp contact configuration $g$ which was originally defined by Ferrari and Canny. The value of $Q(g)$ is defined to be the radius of the largest ball around the origin completely contained in the grasp wrench space $W$, if such a ball exists and $0$ otherwise. This value cannot be computed precisely and a typical approach has hence been to approximate $W$ by a smaller convex hull spanned by finitely many points and which is obtained by approximating the friction cone $F$ at each contact by convex polyhedral cones $F_l$ with $l$ edges.

Polyhedral cone approximations of a friction cone

The resulting grasp quality approximation $Q_l^-(g)\le Q(g)$ has been popular in the grasping community for more than a decade. Our paper "Classical Grasp Quality Evaluation: New Theory and Algorithms", F. T. Pokorny, D. Kragic, IEEE/RSJ IROS, 2013 establishes for the first time error bounds on this approximation and provides a novel approach to also obtain an upper bound $Q^+(g)$ so that $$Q_l^-(g)\le Q(g) \le Q^+(g).$$ This library provides functions to compute $Q_l^-(g)$ and $Q^+(g)$. We use a subgradient-based method to compute $Q^+(g)$ as outlined in the above paper. Additionally, we provide a novel algorithm to speed-up the computation of $Q_l^-(g)$ by first filtering out unstable grasps using a convex optimization procedure. This results in a significant speedup compared to a standard evaluation of $Q_l^-(g)$.

Copyright and Disclaimer

This is a personal homepage. Opinions expressed here or implied by links provided do not represent the official views of KTH Royal Institute of Technology. © Florian T. Pokorny, all rights reserved.