-
Two-headed eye-segmentation approach for biometric identification
Authors:
Wiktor Lazarski,
Maciej Zieba,
Tanguy Jeanneau,
Tobias Zillig,
Christian Brendel
Abstract:
Iris-based identification systems are among the most popular approaches for person identification. Such systems require good-quality segmentation modules that ideally identify the regions for different eye components. This paper introduces the new two-headed architecture, where the eye components and eyelashes are segmented using two separate decoding modules. Moreover, we investigate various trai…
▽ More
Iris-based identification systems are among the most popular approaches for person identification. Such systems require good-quality segmentation modules that ideally identify the regions for different eye components. This paper introduces the new two-headed architecture, where the eye components and eyelashes are segmented using two separate decoding modules. Moreover, we investigate various training scenarios by adopting different training losses. Thanks to the two-headed approach, we were also able to examine the quality of the model with the convex prior, which enforces the convexity of the segmented shapes. We conducted an extensive evaluation of various learning scenarios on real-life conditions high-resolution near-infrared iris images.
△ Less
Submitted 30 September, 2022;
originally announced September 2022.
-
Topologically Protected Transport in Engineered Mechanical Systems
Authors:
Tirth Shah,
Christian Brendel,
Vittorio Peano,
Florian Marquardt
Abstract:
Mechanical vibrations are being harnessed for a variety of purposes and at many length scales, from the macroscopic world down to the nanoscale. The considerable design freedom in mechanical structures allows to engineer new functionalities. In recent years, this has been exploited to generate setups that offer topologically protected transport of vibrational waves, both in the solid state and in…
▽ More
Mechanical vibrations are being harnessed for a variety of purposes and at many length scales, from the macroscopic world down to the nanoscale. The considerable design freedom in mechanical structures allows to engineer new functionalities. In recent years, this has been exploited to generate setups that offer topologically protected transport of vibrational waves, both in the solid state and in fluids. Borrowing concepts from electronic physics and being cross-fertilized by concurrent studies for cold atoms and electromagnetic waves, this field of topological transport in engineered mechanical systems offers a rich variety of phenomena and platforms. In this review, we provide a unifying overview of the various ideas employed in this area, summarize the different approaches and experimental implementations, and comment on the challenges as well as the prospects.
△ Less
Submitted 30 June, 2022; v1 submitted 24 June, 2022;
originally announced June 2022.
-
An Explainable AI System for the Diagnosis of High Dimensional Biomedical Data
Authors:
Alfred Ultsch,
Jörg Hoffmann,
Maximilian Röhnert,
Malte Von Bonin,
Uta Oelschlägel,
Cornelia Brendel,
Michael C. Thrun
Abstract:
Typical state of the art flow cytometry data samples consists of measures of more than 100.000 cells in 10 or more features. AI systems are able to diagnose such data with almost the same accuracy as human experts. However, there is one central challenge in such systems: their decisions have far-reaching consequences for the health and life of people, and therefore, the decisions of AI systems nee…
▽ More
Typical state of the art flow cytometry data samples consists of measures of more than 100.000 cells in 10 or more features. AI systems are able to diagnose such data with almost the same accuracy as human experts. However, there is one central challenge in such systems: their decisions have far-reaching consequences for the health and life of people, and therefore, the decisions of AI systems need to be understandable and justifiable by humans. In this work, we present a novel explainable AI method, called ALPODS, which is able to classify (diagnose) cases based on clusters, i.e., subpopulations, in the high-dimensional data. ALPODS is able to explain its decisions in a form that is understandable for human experts. For the identified subpopulations, fuzzy reasoning rules expressed in the typical language of domain experts are generated. A visualization method based on these rules allows human experts to understand the reasoning used by the AI system. A comparison to a selection of state of the art explainable AI systems shows that ALPODS operates efficiently on known benchmark data and also on everyday routine case data.
△ Less
Submitted 1 March, 2022; v1 submitted 5 July, 2021;
originally announced July 2021.
-
Topological phonon transport in an optomechanical system
Authors:
Hengjiang Ren,
Tirth Shah,
Hannes Pfeifer,
Christian Brendel,
Vittorio Peano,
Florian Marquardt,
Oskar Painter
Abstract:
Recent advances in cavity-optomechanics have now made it possible to use light not just as a passive measuring device of mechanical motion, but also to manipulate the motion of mechanical objects down to the level of individual quanta of vibrations (phonons). At the same time, microfabrication techniques have enabled small-scale optomechanical circuits capable of on-chip manipulation of mechanical…
▽ More
Recent advances in cavity-optomechanics have now made it possible to use light not just as a passive measuring device of mechanical motion, but also to manipulate the motion of mechanical objects down to the level of individual quanta of vibrations (phonons). At the same time, microfabrication techniques have enabled small-scale optomechanical circuits capable of on-chip manipulation of mechanical and optical signals. Building on these developments, theoretical proposals have shown that larger scale optomechanical arrays can be used to modify the propagation of phonons, realizing a form of topologically protected phonon transport. Here, we report the observation of topological phonon transport within a multiscale optomechanical crystal structure consisting of an array of over $800$ cavity-optomechanical elements. Using sensitive, spatially resolved optical read-out we detect thermal phonons in a $0.325-0.34$GHz band traveling along a topological edge channel, with substantial reduction in backscattering. This represents an important step from the pioneering macroscopic mechanical systems work towards topological phononic systems at the nanoscale, where hypersonic frequency ($\gtrsim$GHz) acoustic wave circuits consisting of robust delay lines and non-reciprocal elements may be implemented. Owing to the broadband character of the topological channels, the control of the flow of heat-carrying phonons, albeit at cryogenic temperatures, may also be envisioned.
△ Less
Submitted 13 September, 2020;
originally announced September 2020.
-
Snowflake Topological Insulator for Sound Waves
Authors:
Christian Brendel,
Vittorio Peano,
Oskar Painter,
Florian Marquardt
Abstract:
We show how the snowflake phononic crystal structure, which has been realized experimentally recently, can be turned into a topological insulator for sound waves. This idea, based purely on simple geometrical modifications, could be readily implemented on the nanoscale.
We show how the snowflake phononic crystal structure, which has been realized experimentally recently, can be turned into a topological insulator for sound waves. This idea, based purely on simple geometrical modifications, could be readily implemented on the nanoscale.
△ Less
Submitted 23 January, 2017;
originally announced January 2017.
-
Pseudomagnetic fields for sound at the nanoscale
Authors:
Christian Brendel,
Vittorio Peano,
Oskar Painter,
Florian Marquardt
Abstract:
There is a growing effort in creating chiral transport of sound waves. However, most approaches so far are confined to the macroscopic scale. Here, we propose a new approach suitable to the nanoscale which is based on pseudomagnetic fields. These fields are the analogon for sound of the pseudomagnetic field for electrons in strained graphene. In our proposal, they are created by simple geometrical…
▽ More
There is a growing effort in creating chiral transport of sound waves. However, most approaches so far are confined to the macroscopic scale. Here, we propose a new approach suitable to the nanoscale which is based on pseudomagnetic fields. These fields are the analogon for sound of the pseudomagnetic field for electrons in strained graphene. In our proposal, they are created by simple geometrical modifications of an existing and experimentally proven phononic crystal design, the snowflake crystal. This platform is robust, scalable, and well-suited for a variety of excitation and readout mechanisms, among them optomechanical approaches.
△ Less
Submitted 14 July, 2016;
originally announced July 2016.
-
Topological phase transitions and chiral inelastic transport induced by the squeezing of light
Authors:
Vittorio Peano,
Martin Houde,
Christian Brendel,
Florian Marquardt,
Aashish A. Clerk
Abstract:
We show how the squeezing of light can lead to the formation of topological states. Such states are characterized by non-trivial Chern numbers, and exhibit protected edge modes which give rise to chiral elastic and inelastic photon transport. These topological bosonic states are not equivalent to their fermionic (topological superconductor) counterparts and cannot be mapped by a local transformati…
▽ More
We show how the squeezing of light can lead to the formation of topological states. Such states are characterized by non-trivial Chern numbers, and exhibit protected edge modes which give rise to chiral elastic and inelastic photon transport. These topological bosonic states are not equivalent to their fermionic (topological superconductor) counterparts and cannot be mapped by a local transformation onto topological states found in particle-conserving models. They thus represent a new type of topological system. We study this physics in detail in the case of a Kagome lattice model, and discuss possible realizations using nonlinear photonic crystals or superconducting circuits.
△ Less
Submitted 6 August, 2015;
originally announced August 2015.
-
Pattern phase diagram for 2D arrays of coupled limit-cycle oscillators
Authors:
Roland Lauter,
Christian Brendel,
Steven J. M. Habraken,
Florian Marquardt
Abstract:
Arrays of coupled limit-cycle oscillators represent a paradigmatic example for studying synchronization and pattern formation. They are also of direct relevance in the context of currently emerging experiments on nano- and optomechanical oscillator arrays. We find that the full dynamical equations for the phase dynamics of such an array go beyond previously studied Kuramoto-type equations. We anal…
▽ More
Arrays of coupled limit-cycle oscillators represent a paradigmatic example for studying synchronization and pattern formation. They are also of direct relevance in the context of currently emerging experiments on nano- and optomechanical oscillator arrays. We find that the full dynamical equations for the phase dynamics of such an array go beyond previously studied Kuramoto-type equations. We analyze the evolution of the phase field in a two-dimensional array and obtain a "phase diagram" for the resulting stationary and non-stationary patterns. The possible observation in optomechanical arrays is discussed briefly.
△ Less
Submitted 7 January, 2015;
originally announced January 2015.
-
Topological Phases of Sound and Light
Authors:
V. Peano,
C. Brendel,
M. Schmidt,
F. Marquardt
Abstract:
Topological states of matter are particularly robust, since they exploit global features insensitive to local perturbations. In this work, we describe how to create a Chern insulator of phonons in the solid state. The proposed implementation is based on a simple setting, a dielectric slab with a suitable pattern of holes. Its topological properties can be wholly tuned in-situ by adjusting the ampl…
▽ More
Topological states of matter are particularly robust, since they exploit global features insensitive to local perturbations. In this work, we describe how to create a Chern insulator of phonons in the solid state. The proposed implementation is based on a simple setting, a dielectric slab with a suitable pattern of holes. Its topological properties can be wholly tuned in-situ by adjusting the amplitude and frequency of a driving laser that controls the optomechanical interaction between light and sound. The resulting chiral, topologically protected phonon transport along the edges can be probed completely optically. Moreover, we identify a regime of strong mixing between photon and phonon excitations, which gives rise to a large set of different topological phases. This would be an example of a Chern insulator produced from the interaction between two physically very different particle species, photons and phonons.
△ Less
Submitted 28 August, 2015; v1 submitted 18 September, 2014;
originally announced September 2014.
