Kechen Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Kechen Zhang has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Alexandre Pouget
Department of Basic Neuroscience, University of Geneva, Geneva, Switzerland
1 shared publication - Brian S. Robinson
Johns Hopkins University Applied Physics Laboratory, Laurel, Maryland, UNITED STATES
1 shared publication - Caixia Wang
School of International Economics, China Foreign Affairs University, Beijing, China
1 shared publication - Chenhao Bao
Department of Biomedical Engineering, Johns Hopkins University, Baltimore, MD, 21205, USA
1 shared publication Danilo Symonette
Johns Hopkins University Applied Physics Laboratory, Laurel, Maryland, UNITED STATES
1 shared publicationErik C. Johnson
Johns Hopkins University Applied Physics Laboratory, Laurel, Maryland, UNITED STATES
1 shared publication
Latest Publications
- Open Access
A canonical input to the head direction compass carries predominantly eye movement signals
The mammalian head direction (HD) system is thought to update its internal compass using angular head velocity (AHV) signals relayed from the brainstem. We tested this canonical interpretation by recording from the nucleus prepositus hypoglossi (NPH), a proposed AHV relay known to be important for accurate HD tracking, while simultaneously measuring eye and head movements in mice. Neurons classified as AHV-tuned instead predominantly encoded eye movements: most showed saccade-related bursts or pauses, and apparent AHV tuning was largely explained by eye rather than head motion. Optotagging showed that even NPH neurons projecting directly to the HD circuit predominantly encoded eye movements. Empirically constrained population models reconstructed eye position and gaze direction, but head-direction readouts failed when eye and head motion were dissociated. These findings overturn the canonical view of NPH as a dedicated AHV relay, revealing eye movements as a major, previously unmeasured component of the signals entering the mammalian HD circuit and thereby changing the computational problem that the HD system must solve to maintain a stable representation of heading.
Open AccessA canonical input to the head direction compass carries predominantly eye movement signals
The mammalian head direction (HD) system is thought to update its internal compass using angular head velocity (AHV) signals relayed from the brainstem. We tested this canonical interpretation by recording from the nucleus prepositus hypoglossi (NPH), a proposed AHV relay known to be important for accurate HD tracking, while simultaneously measuring eye and head movements in mice. Neurons classified as AHV-tuned instead predominantly encoded eye movements: most showed saccade-related bursts or pauses, and apparent AHV tuning was largely explained by eye rather than head motion. Optotagging showed that even NPH neurons projecting directly to the HD circuit predominantly encoded eye movements. Empirically constrained population models reconstructed eye position and gaze direction, but head-direction readouts failed when eye and head motion were dissociated. These findings overturn the canonical view of NPH as a dedicated AHV relay, revealing eye movements as a major, previously unmeasured component of the signals entering the mammalian HD circuit and thereby changing the computational problem that the HD system must solve to maintain a stable representation of heading.
- Open Access
Online learning for orientation estimation during translation in an insect ring attractor network
Insect neural systems are a promising source of inspiration for new algorithms for navigation, especially on low size, weight, and power platforms. There have been unprecedented recent neuroscience breakthroughs with Drosophila in behavioral and neural imaging experiments as well as the mapping of detailed connectivity of neural structures. General mechanisms for learning orientation in the central complex (CX) of Drosophila have been investigated previously; however, it is unclear how these underlying mechanisms extend to cases where there is translation through an environment (beyond only rotation), which is critical for navigation in robotic systems. Here, we develop a CX neural connectivity-constrained model that performs sensor fusion, as well as unsupervised learning of visual features for path integration; we demonstrate the viability of this circuit for use in robotic systems in simulated and physical environments. Furthermore, we propose a theoretical understanding of how distributed online unsupervised network weight modification can be leveraged for learning in a trajectory through an environment by minimizing of orientation estimation error. Overall, our results here may enable a new class of CX-derived low power robotic navigation algorithms and lead to testable predictions to inform future neuroscience experiments. Summary: An insect neural connectivity-constrained model performs sensor fusion and online learning for orientation estimation.
Open AccessOnline learning for orientation estimation during translation in an insect ring attractor network
Insect neural systems are a promising source of inspiration for new algorithms for navigation, especially on low size, weight, and power platforms. There have been unprecedented recent neuroscience breakthroughs with Drosophila in behavioral and neural imaging experiments as well as the mapping of detailed connectivity of neural structures. General mechanisms for learning orientation in the central complex (CX) of Drosophila have been investigated previously; however, it is unclear how these underlying mechanisms extend to cases where there is translation through an environment (beyond only rotation), which is critical for navigation in robotic systems. Here, we develop a CX neural connectivity-constrained model that performs sensor fusion, as well as unsupervised learning of visual features for path integration; we demonstrate the viability of this circuit for use in robotic systems in simulated and physical environments. Furthermore, we propose a theoretical understanding of how distributed online unsupervised network weight modification can be leveraged for learning in a trajectory through an environment by minimizing of orientation estimation error. Overall, our results here may enable a new class of CX-derived low power robotic navigation algorithms and lead to testable predictions to inform future neuroscience experiments. Summary: An insect neural connectivity-constrained model performs sensor fusion and online learning for orientation estimation.
- Open Access
- 23 January 2020
- Published by Frontiers Media SA in Frontiers in Computational Neuroscience
- Vol. 13, 96
Equilibrium States and Their Stability in the Head-Direction Ring Network
Head-direction cells have been found in several areas in the mammalian brains. The firing rate of an ideal head-direction cell reaches its peak value only when the animal's head points in a specific direction, and this preferred direction stays the same regardless of spatial location. In this paper we combine mathematical analytical techniques and numerical simulations to fully analyze the equilibrium states of a generic ring attractor network, which is a widely used modeling framework for the head-direction system. Under specific conditions, all solutions of the ring network are bounded, and there exists a Lyapunov function that guarantees the stability of the network for any given inputs, which may come from multiple sources in the biological system, including self-motion information for inertially based updating and landmark information for calibration. We focus on the first few terms of the Fourier series of the ring network to explicitly solve for all possible equilibrium states, followed by a stability analysis based on small perturbations. In particular, these equilibrium states include the standard single-peaked activity pattern as well as double-peaked activity pattern, whose existence is unknown but has testable experimental implications. To our surprise, we have also found an asymmetric equilibrium activity profile even when the network connectivity is strictly symmetric. Finally we examine how these different equilibrium solutions depend on the network parameters and obtain the phase diagrams in the parameter space of the ring network.
- 23 January 2020
- Published by Frontiers Media SA in Frontiers in Computational Neuroscience
- Vol. 13, 96
Open AccessEquilibrium States and Their Stability in the Head-Direction Ring Network
Head-direction cells have been found in several areas in the mammalian brains. The firing rate of an ideal head-direction cell reaches its peak value only when the animal's head points in a specific direction, and this preferred direction stays the same regardless of spatial location. In this paper we combine mathematical analytical techniques and numerical simulations to fully analyze the equilibrium states of a generic ring attractor network, which is a widely used modeling framework for the head-direction system. Under specific conditions, all solutions of the ring network are bounded, and there exists a Lyapunov function that guarantees the stability of the network for any given inputs, which may come from multiple sources in the biological system, including self-motion information for inertially based updating and landmark information for calibration. We focus on the first few terms of the Fourier series of the ring network to explicitly solve for all possible equilibrium states, followed by a stability analysis based on small perturbations. In particular, these equilibrium states include the standard single-peaked activity pattern as well as double-peaked activity pattern, whose existence is unknown but has testable experimental implications. To our surprise, we have also found an asymmetric equilibrium activity profile even when the network connectivity is strictly symmetric. Finally we examine how these different equilibrium solutions depend on the network parameters and obtain the phase diagrams in the parameter space of the ring network.
- Open Access
- 19 August 1999
- Published by MIT Press
- p. 55-84
Statistically Efficient Estimation Using Population Coding
- 19 August 1999
- Published by MIT Press
- p. 55-84
Open AccessStatistically Efficient Estimation Using Population Coding