A5028991390- Retinal Development and Disorders
- Photoreceptor and optogenetics research
- Neural dynamics and brain function
- Advanced Fluorescence Microscopy Techniques
- Neuroscience and Neural Engineering
- Video Surveillance and Tracking Methods
- Nitric Oxide and Endothelin Effects
- Anomaly Detection Techniques and Applications
- Radar Systems and Signal Processing
- Circadian rhythm and melatonin
- Fault Detection and Control Systems
- Visual perception and processing mechanisms
- Neurological Disease Mechanisms and Treatments
- CCD and CMOS Imaging Sensors
- Artificial Intelligence in Healthcare
- Neurobiology and Insect Physiology Research
- Advanced SAR Imaging Techniques
- Video Analysis and Summarization
- Retinal Imaging and Analysis
- Cell Image Analysis Techniques
- Underwater Acoustics Research
- Advanced Neural Network Applications
- Neuroscience, Education and Cognitive Function
- Intelligent Tutoring Systems and Adaptive Learning
- Magnetic Bearings and Levitation Dynamics
University of Tübingen
2018-2025
Bernstein Center for Computational Neuroscience Tübingen
2018-2024
Inserm
2024
Sorbonne Université
2024
Institut de la Vision
2024
Centre National de la Recherche Scientifique
2024
Nanjing Polytechnic Institute
2023
Naval Aeronautical and Astronautical University
2011-2020
Shaanxi University of Science and Technology
2018
Southeast University
2011
Visual neuroscientists require accurate control of visual stimulation. However, few stimulator solutions simultaneously offer high spatio-temporal resolution and free over the spectra light sources, because they rely on off-the-shelf technology developed for human trichromatic vision. Importantly, consumer displays fail to drive UV-shifted short wavelength-sensitive photoreceptors, which strongly contribute behaviour in many animals, including mice, zebrafish fruit flies. Moreover,...
The task of crowd counting in varying density scenes is an extremely difficult challenge due to large scale variations. In this paper, we propose a novel dual path multi-scale fusion network architecture with attention mechanism named SFANet that can perform accurate count estimation as well present high-resolution maps for highly congested scenes. proposed contains two main components: VGG backbone convolutional neural (CNN) the front-end feature map extractor and networks back-end generate...
Abstract The retina decomposes visual stimuli into parallel channels that encode different features of the environment. Central to this computation is synaptic processing in a dense layer neuropil, so-called inner plexiform (IPL). Here, types bipolar cells stratifying at distinct depths relay excitatory feedforward drive from photoreceptors amacrine and ganglion cells. Current experimental techniques for studying IPL do not allow imaging entire simultaneously intact tissue. we extend...
Neuromodulators have major influences on the regulation of neural circuit activity across nervous system. Nitric oxide (NO) has been shown to be a prominent neuromodulator in many circuits and extensively studied retina. Here, it associated with light adaptation, gain control, gap junctional coupling, but its effect retinal output, specifically different types ganglion cells (RGCs), is still poorly understood. In this study, we used two-photon Ca 2+ imaging multi-electrode array (MEA)...
Although 2-D canonical correlation analysis (2DCCA) has been proposed to reduce the computational complexity while reserving local data structure of image, learned variables 2DCCA are linear combination all original variables, which makes it hard interpret solutions and might have less generality. In this paper, we propose a sparse (S2DCCA) solve drawbacks method apply image feature extraction. The basic idea S2DCCA is impose two lasso penalties on objective function obtain sets projection...
More and more football fans prefer to watch highlights because it takes too much time the entire game. At same time, many TV stations sports companies need store retrieve fragments of different events in matches. The traditional way is analyze cut match manually. This method consuming process largely depends on personal choice, which results unsecured accuracy. Therefore, it's necessary develop a fast accurate system for automatic summarization analysis matches video. Automatic event...
Neuromodulators have major influences on the regulation of neural circuit activity across nervous system. Nitric oxide (NO) has been shown to be a prominent neuromodulator in many circuits and extensively studied retina. Here, it associated with light adaptation, gain control, gap junctional coupling, but its effect retinal output, specifically different types ganglion cells (RGCs), is still poorly understood. In this study, we used two-photon Ca 2+ imaging multi-electrode array (MEA)...
Abstract The retina decomposes visual stimuli into parallel channels that encode different features of the environment. Central to this computation is synaptic processing in a dense and thick layer neuropil, so-called inner plexiform (IPL). Here, types bipolar cells stratifying at distinct depths relay excitatory feedforward drive from photoreceptors amacrine ganglion cells. Current experimental techniques for studying IPL do not allow imaging entire simultaneously intact tissue. we extend...
Variability, stochastic or otherwise, is a central feature of neural activity. Yet the means by which estimates variation and uncertainty are derived from noisy observations activity often heuristic, with more weight given to numerical convenience than statistical rigour. For two-photon imaging data, composed fundamentally probabilistic streams photon detections, problem particularly acute. Here, we present pipeline for inference analysis using Gaussian Process regression, applied recordings...
Summary Pressures for survival drive sensory circuit adaption to a species’ habitat, making it essential statistically characterise natural scenes. Mice, prominent visual system model, are dichromatic with enhanced sensitivity green and UV. Their environment, however, is rarely considered. Here, we built UV-green camera record footage from mouse habitats. We found chromatic contrast greatly diverge in the upper but not lower field, an environmental difference that may underlie superior...
Based on modified rank test, this paper proposes two new nonparametric detectors by using the K-distributed sea clutter model. These are Modified-Rank-based Generalized Sign detector (MRGS) and Mann-Whitney (MRMW). The achieved designing inverse-normal-score functions of statistics. They CFAR (Constant False Alarm Rate) for data with continuous CDF (Cumulative Distribution Function) PDF (Probability Density bing symmetric about zero. performance is analyzed under Swerling II target model...
Abstract Visual neuroscientists require accurate control of visual stimulation. However, few stimulator solutions simultaneously offer high spatio-temporal resolution and free over the spectra light sources, because they rely on off-the-shelf technology developed for human trichromatic vision. Importantly, consumer displays fail to drive UV-shifted short wavelength-sensitive photoreceptors, which strongly contribute behaviour in many animals, including mice, zebrafish fruit flies. Moreover,...
This paper mainly proposed to identify and classify the medical data of cervical cancer using neural network models such as SVM, FNN, KNN so on.The computer recognition algorithm can overcome deficiency that artificial identification tends be affected by cognitive ability, subjective experience fatigue degree.The model trained under various networks with experts' manual marked will obtain a more precise result in process data.
The micro-motion of a subject induces Doppler frequency modulations around the carrier reflected sensor signals. Recently, it has been proved that sea clutter is significantly variable and surface subjects have their micro-motions influenced by state. micro-Doppler (m-D) signatures can describe refined motion characteristics target. micromotion signal model target established in this paper based on length observation time. Finally, properties are analyzed using real radar data, i.e., X-band...
A novel long-time coherent integration method, i.e., scaling processing and Radon-Fourier transform (SPRFT), is proposed to detect the low-observable manuevering target, which can compensate range Doppler migrations simultaneously. The target signal be well accumulated as an obvious peak in centroid frequency-chirp rate (CFCR) domain by simple FT calculations. It shown that SPRFT a generalized form of RFT Lv's distribution (LVD), includes instantaneous autocorrelation function, process,...
Videos are dominant on the Internet. Current systems to process large-scale videos suboptimal due following reasons: (1) machine learning modules such as feature extractors and classifiers generate huge intermediate data place heavy burden storage network, (2) task scheduling is explicit; manually configuring cluster tedious inefficient. In this work, we propose Elastic Streaming Sequential Processing system (ESSP) that supports automatic scheduling; multiple components automatically...
Abstract The visual system of any animal species is adapted to its ecological niche. Thus, investigating function and behavior using naturalistic stimuli holds significant potential. In mice, these adaptations include a field view ∼280° cone opsins sensitive UV green wavelengths. Such adaptations, however, cannot be probed with standard consumer displays. To present we built hemispheric dome setup, enabling the controlled projection wide-field movies UV-green spectral content. For our...