Holger Finger

University Medical Center Hamburg-Eppendorf

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Peter König

University Medical Center Hamburg-Eppendorf

Christian Gerloff

University Medical Center Hamburg-Eppendorf

Flemming Brieger

Helmholtz-Institute Ulm

Gordon Pipa

Osnabrück University

Sonja Füllhase

Leon René Sütfeld

RISE Research Institutes of Sweden

Moritz Köster

University of Regensburg

Caspar Goeke

Osnabrück University

Thomas Gruber

Osnabrück University

Andreas K. Engel

Universität Hamburg

Cooperative Institutions

Osnabrück University

Universität Hamburg

University Medical Center Hamburg-Eppendorf

Klinik und Poliklinik für Neurologie

Max Planck Institute for Human Cognitive and Brain Sciences

Max Planck Society

German Center for Neurodegenerative Diseases

Charité - Universitätsmedizin Berlin

Université Claude Bernard Lyon 1

Hospices Civils de Lyon

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Webcam eye tracking close to laboratory standards: Comparing a new webcam-based system and the EyeLink 1000

Behavior Research Methods (2023)

Tobiasz Kaduk Caspar Goeke Holger Finger Peter König

Abstract This paper aims to compare a new webcam-based eye-tracking system, integrated into the Labvanced platform for online experiments, to a “gold standard” lab-based eye tracker (EyeLink 1000 - SR Research). Specifically, we simultaneously recorded data with both eye trackers in five different tasks, analyzing their real-time performance. These tasks were a subset of a standardized test battery for eye trackers, including a Large Grid task, Smooth Pursuit eye movements, viewing natural images, and two Head Movements tasks (roll, yaw). The results show that the webcam-based system achieved an overall accuracy of 1.4°, and a precision of 1.1° (standard deviation (SD) across subjects), an error of about 0.5° larger than the EyeLink system. Interestingly, both accuracy (1.3°) and precision (0.9°) were slightly better for centrally presented targets, the region of interest in many psychophysical experiments. Remarkably, the correlation of raw gaze samples between the EyeLink and webcam-based was at about 90% for the Large Grid task and about 80% for Free View and Smooth Pursuit. Overall, these results put the performance of the webcam-based system roughly on par with mobile eye-tracking devices (Ehinger et al. PeerJ , 7 , e7086, 2019; Tonsen et al., 2020) and demonstrate substantial improvement compared to existing webcam eye-tracking solutions (Papoutsaki et al., 2017).

BitTorrent tracker

Tracking (education)

10.3758/s13428-023-02237-8

Cite

Citations (17)

Probing Neural Networks for Dynamic Switches of Communication Pathways

bioRxiv (Cold Spring Harbor Laboratory) (2019)

Holger Finger Richard Gast Christian Gerloff Andreas K. Engel Peter König

Abstract Dynamic communication and routing play important roles in the human brain to facilitate 2exibility in task solving and thought processes. Here, we present a network perturbation methodology that allows to investigate dynamic switching between different network pathways based on phase offsets between two external oscillatory drivers. We apply this method in a computational model of the human connectome with delay-coupled neural masses. To analyze dynamic switching of pathways, we define four new metrics that measure dynamic network response properties for pairs of stimulated nodes. Evaluating these metrics for all network pathways, we found a broad spectrum of pathways with distinct dynamic properties and switching behaviors. Specifically, we found that 60.1% of node pairs can switch their communication from one pathway to another depending on their phase offsets. This indicates that phase offsets and coupling delays play an important computational role for the dynamic switching between communication pathways in the brain.

Adaptive routing

Dynamic network analysis

10.1101/523522

Cite

Citations (0)

Approaches and databases for online calibration of binaural sound localization for robotic heads

2011 IEEE/RSJ International Conference on Intelligent Robots and Systems (2010)

Holger Finger Shih‐Chii Liu Paul Ruvolo Javier R. Movellan

In this paper, we evaluate adaptive sound localization algorithms for robotic heads. To this end we built a 3 degree-of-freedom head with two microphones encased in artificial pinnae (outer ears). The geometry of the head and pinnae induce temporal differences in the sound recorded at each microphone. These differences change with the frequency of the sound, location of the sound, and orientation of the robot in a complex manner. To learn the relationship between these auditory differences and the location of a sound source, we applied machine learning methods to a database of different audio source locations and robot head orientations. Our approach achieves a mean error of 2.5 degrees for azimuth and 11 degrees for elevation for estimating the position of an audio source. The impressive results highlight the benefits of a two-stage regression model to make use of the properties of the artificial pinnae for elevation estimation. In this work, the algorithms were trained using ground truth data provided by a motion capture system. We are currently generalizing the approach so that the training signal is provided online based on a real-time face detection and speech detection system.

Head-related transfer function

Ground truth

Position (finance)

RANSAC

10.1109/iros.2010.5650515

Cite

Citations (18)

Adaptive Blending Units: Trainable Activation Functions for Deep Neural Networks

arXiv (Cornell University) (2018)

Leon René Sütfeld Flemming Brieger Holger Finger Sonja Füllhase Gordon Pipa

The most widely used activation functions in current deep feed-forward neural networks are rectified linear units (ReLU), and many alternatives have been successfully applied, as well. However, none of the alternatives have managed to consistently outperform the rest and there is no unified theory connecting properties of the task and network with properties of activation functions for most efficient training. A possible solution is to have the network learn its preferred activation functions. In this work, we introduce Adaptive Blending Units (ABUs), a trainable linear combination of a set of activation functions. Since ABUs learn the shape, as well as the overall scaling of the activation function, we also analyze the effects of adaptive scaling in common activation functions. We experimentally demonstrate advantages of both adaptive scaling and ABUs over common activation functions across a set of systematically varied network specifications. We further show that adaptive scaling works by mitigating covariate shifts during training, and that the observed advantages in performance of ABUs likewise rely largely on the activation function's ability to adapt over the course of training.

Activation function

Source

Cite

Citations (9)