National Repository of Grey Literature 10 records found  Search took 0.02 seconds. 
A model inspection/visualization tool for brain simulation framework.
Ježek, Filip ; Antolík, Ján (advisor) ; Pešková, Klára (referee)
Simulations of biological neural networks are an important tool for under- standing how the brain processes information. Mozaik is a workflow frame- work that allows such simulations to be created, run and analyzed. Currently, however, there is no easy way to visualize in it either the network or the data structures created by individual analyses. This work is a web application that allows users to visualize networks and data structures stored in data- stores for individual simulations. The given visualizations can be searched using complex queries, interactively examined and compared with each other. This makes the work of researchers working with the Mozaik framework much more efficient.
Learning V1 targeting optogenetic stimulation protocol for inducing visual perception
Parada, Jakub ; Antolík, Ján (advisor) ; Korvasová, Karolína (referee)
The Optogenetic stimulation of neurons in the primary visual cortex (V1) is a novel and promising technique for vision restoration for people with acquired blindness. One of the challenges of such a technique is finding artificial stimuli which invoke desired cortical activities. This thesis explores whether neural networks and deep learning can be used for reverse engineering artificial stimuli patterns for optogenetic cortical implant prosthesis (LED) from cortical activity pattern recordings, assuming that similar cortical activity recordings are caused by similar visual stimuli. Various DNN architectures outperforming baseline solutions in stimulus reconstruction will be explored. Loss functions such as MSE and Structural similarity (SSIM) will be used. Questions such as if loss of information in the high-frequency domain of the reconstructed stimuli negatively affects correspondence between the desired cortical activity and the activity elicited by artificial stimuli patterns will be investigated. MSE evaluation metric will be used to determine the degree of similarity between the two types of cortical activities. Due to the limited availability of biological data, we use a model of V1 combined with a model of optogenetic cortical prosthesis (LED) and stimulation developed by Antolík et al. [2021] to...
Characterizing computations in a model of biological vision using deep-neural-network approaches.
Nepožitek, David ; Antolík, Ján (advisor) ; Hoksza, David (referee)
In this thesis, we examine two kinds of models of the primary visual cor- tex: a deep neural network for system identification and a spiking model of a cat's primary visual cortex. Further progress in modelling visual sys- tems can help us comprehend the brain's inner workings in greater detail; moreover, it can help to develop better visual prosthesis or further improve models that handle visual inputs, such as those used for object classification. We employ the state-of-the-art deep neural network to predict the responses of the spiking model when presented with natural stimuli. We demonstrate that by tuning the hyperparameters, the deep neural network explains ap- proximately 85% of the explainable variance observed in the responses of the spiking model. That is significantly more accurate than predictions of real neural responses, suggesting that real neurons possess certain charac- teristics not captured in the spiking model. However, we also argue that the network would not be capable of perfect predictions even when a large amount of data is provided. We show that the network encounters notable difficulties in modelling neurons with high noise and precisely predicting high firing rates. Furthermore, we analyse the network's representations by phase, orientation and size tuning. We illustrate that the...
Decoding visual stimuli from cortical activity
Vašek, Vojtěch ; Antolík, Ján (advisor) ; Šikudová, Elena (referee)
This thesis aims to develop a machine learning model that can decode stimu- lus images from cortical activity in the primary visual cortex (V1) to understand the relationship between V1 activity and visual perception. The limited avai- lability of biological data makes it necessary to use the spiking neural network model of V1 to generate the underlying training data. Machine learning tech- niques, particularly neural networks, will be explored to generate high-quality stimulus images. Standard loss functions, as well as discriminator loss from GAN networks training, will be used to train the decoding models. Linear regression models will be used baseline. The research questions to be addressed include the best decoding approach, the impact of the number of neurons recorded or stimuli presented, the loss of information in high frequencies domain and the effect of intrinsic noise in neural responses on reconstructing visual stimuli. This thesis proposes a trainable convolutional network, which outperforms other baseline models such as linear regression. We observe that the loss function producing the best results is the MSSSIM. However, the intrinsic noise in neural respon- ses limits the reconstruction, and only low frequencies are being reconstructed. The size of the dataset and the number of cortical...
Predicting accuracy in Multiple Object Tracking tasks from trajectory statistics
Chembrolu, Surya Prakash ; Děchtěrenko, Filip (advisor) ; Antolík, Ján (referee)
Title: Predicting accuracy in Multiple Object Tracking tasks from trajectory statistics Author: Surya Prakash Chembrolu Department: Department of Software and Computer Science Education Supervisor: Mgr. Filip Děchtěrenko, Ph.D., Department of Software and Com- puter Science Education Abstract: Cognitive science is an interdisciplinary area covering neuroscience, psychology, linguistics, philosophy, and computer science. Computer science and cognitive science mutually benefit from each other because computer science is very helpful to design and perform experiments in order to understand how the brain works likewise research output from cognitive science can lead to new con- cepts and models in artificial intelligence. Within cognitive science, Multiple Object Tracking (MOT) paradigm is used to study visual attention. In MOT experiments, participants are required to keep track of some moving objects in parallel. In this study, a data-driven approach is taken in order to explain the tracking performance of the subjects taking part in MOT experiments. The stimuli in MOT known as trajectories or tracks presented in previous studies were taken and the difficulty of those trajectories is quantified based on trajec- tory statistics. Then a model is created to explain tracking performance and this model is tested...
Rotation-equivariant convolutional neural network for design of visual prosthetic stimulation protocol
Picek, Martin ; Antolík, Ján (advisor) ; Pilát, Martin (referee)
Neighboring neurons in the primary visual cortex (V1), the first cortical area pro- cessing visual information, are selective to stimuli presented in neighboring positions of the visual field with a specific edge orientation. In this way, they form the so-called retinotopic and orientation maps of V1. Due to the absence of high-resolution cortical stimulation devices, vision restoration through prosthetic implants in V1 has not yet taken advantage of the orientation maps. However, the availability of cortical implants with stimulation resolution high enough to target separate orientation columns can be anticipated soon. Since other stimulus features are also encoded in the cortex, such as color, size, or phase, but cannot be reliably engaged even by high-resolution stimulation, in this thesis, we ask the question of how well can visual stimuli be encoded in V1 if only orientation and position preference is known. To address this question, we propose a deep neural network (DNN) providing a scalar neural activity descriptor for any targeted cortical location and multiple different orientations. This is achieved by employing a rotation-equivariant convolutional neural network (reCNN) with the last layer having only one channel for each orientation, returning the desired three-dimensional feature tensor. A...
Finding invariances in sensory coding through gradient methods.
Kovács, Peter ; Antolík, Ján (advisor) ; Šikudová, Elena (referee)
The key to understanding vision is to acquire insight into the sensory coding of indi- vidual neurons. To this end, major advances were done over the past 50 years in fitting models to neural data to identify the mapping from sensory space to neural responses. Especially the advance of DNNs in neuroscience allowed for model fits with excellent predictive power. However, such advanced neural models are complex, and their poor in- terpretability has so far hindered deeper understanding of the principles of visual coding. To address this issue, a recent study proposed a method which identifies the stimulus that activates the neuron the most. However, the sensory coding of highly non-linear neurons, which are abundant already at the earliest stages of visual processing, is too complex for a single stimulus to sufficiently characterize it. A more robust way to char- acterize this coding is through identifying the input sub-space within which the neuron is activated identically - i.e. finding invariances of the neuron's sensory representation. In this thesis, a novel approach for finding such invariant stimuli is proposed. The proposed technique is based on a generator neural network, which maps Gaussian noise from latent space to a stimulus set which equally activates a given neuron. The method demonstrated the...
Predikce rychlosti a absolutni rychlosti pohybu z lidských intrakraniálních EEG dat pomocí hlubokých neuronových sítí.
Vystrčilová, Michaela ; Antolík, Ján (advisor) ; Pilát, Martin (referee)
Our brain controls the processes of the body including movement. In this thesis, we try to understand how the information about hand movement is encoded into the brain's electrical activity and how this activity can be used to predict the velocity and absolute velocity of hand movements. Using a well-established deep neural network architecture for EEG decoding - the Deep4Net - we predict hand movement velocity and absolute velocity from intracranial EEG signals. While reaching the expected performance level, we determine the influence of different frequency bands on the network's prediction. We find that modulations in the high-gamma frequency band are less informative than expected based on previous studies. We also identify two architectural modifications which lead to higher performances. 1. the removal of max-pooling layers in the architecture leads to significantly higher correlations. 2. the non-uniform receptive field of the network is a potential drawback making the network biased towards less relevant information. 1
Deep-learning architectures for analysing population neural data
Houška, Petr ; Antolík, Ján (advisor) ; Pilát, Martin (referee)
Accurate models of visual system are key for understanding how our brains process visual information. In recent years, deep neural networks (DNN) have been rapidly gaining traction in this domain. However, only few studies attempted to incorporate known anatomical properties of visual system into standard DNN architectures adapted from the general machine learning field, to improve their interpretability and performance on visual data. In this thesis, we optimize a recent biologically inspired deep learning architecture designed for analysis of population data recorded from mammalian primary visual cortex when presented with natural images as stimuli. We reimplement this prior modeling in existing neuroscience focused deep learning framework NDN3 and assess it in terms of stability and sensitivity to hyperparameters and architecture fine-tuning. We proceed to extend the model with components of various DNN models, analysing novel combinations and techniques from classical computer vision deep learning, comparing their effectiveness against the bio-inspired components. We were able to identify modifications that greatly increase the stability of the model while securing moderate improvement in overall performance. Furthermore, we document the importance of small hyperparameters adjustments versus...

Interested in being notified about new results for this query?
Subscribe to the RSS feed.