A neurocomputational model of figure-ground discrimination and target tracking

Artificial fly visual joint perception neural network inspired by multiple-regional collision detection

The biological visual system includes multiple types of motion sensitive neurons which preferentially respond to specific perceptual regions. However, it still keeps open how to borrow such neurons to construct bio-inspired computational models for multiple-regional collision detection. To fill this gap, this work proposes a visual joint perception neural network with two subnetworks - presynaptic and postsynaptic neural networks, inspired by the preferentialperception characteristics of three horizontal and vertical motion sensitive neurons. Related to the neural network and three hazard detection mechanisms, an artificial fly visual synthesized collision detection model for multiple-regional collision detection is originally developed to monitor possible danger occurrence in the case where one or more moving objects appear in the whole field of view. The experiments can clearly draw two conclusions: (i) the acquired neural network can effectively display the characteristics of visual movement, and (ii) the collision detection model, which outperforms the compared models, can effectively perform multiple-regional collision detection at a high success rate, and only takes about 0.24s to complete the process of collision detection for each virtual or actual image frame with resolution 110×60.

Lateral inhibition pyramidal neural network for image classification

The human visual system is one of the most fascinating and complex mechanisms of the central nervous system that enables our capacity to see. It is through the visual system that we are able to accomplish from the most simple task such as object recognition to the most complex visual interpretation, understanding and perception. Inspired by this sophisticated system, two models based on the properties of the human visual system are proposed. These models are designed based on the concepts of receptive and inhibitory fields. The first model is a pyramidal neural network with lateral inhibition, called lateral inhibition pyramidal neural network. The second proposed model is a supervised image segmentation system, called segmentation and classification based on receptive fields. This work shows that the combination of these two models is beneficial, and the results obtained are better than that of other state-of-the-art methods.

Noise destroys feedback enhanced figure-ground segmentation but not feedforward figure-ground segmentation

Figure-ground (FG) segmentation is the separation of visual information into background and foreground objects. In the visual cortex, FG responses are observed in the late stimulus response period, when neurons fire in tonic mode, and are accompanied by a switch in cortical state. When such a switch does not occur, FG segmentation fails. Currently, it is not known what happens in the brain on such occasions. A biologically plausible feedforward spiking neuron model was previously devised that performed FG segmentation successfully. After incorporating feedback the FG signal was enhanced, which was accompanied by a change in spiking regime. In a feedforward model neurons respond in a bursting mode whereas in the feedback model neurons fired in tonic mode. It is known that bursts can overcome noise, while tonic firing appears to be much more sensitive to noise. In the present study, we try to elucidate how the presence of noise can impair FG segmentation, and to what extent the feedforward and feedback pathways can overcome noise. We show that noise specifically destroys the feedback enhanced FG segmentation and leaves the feedforward FG segmentation largely intact. Our results predict that noise produces failure in FG perception.

Feedback enhances feedforward figure-ground segmentation by changing firing mode

In the visual cortex, feedback projections are conjectured to be crucial in figure-ground segregation. However, the precise function of feedback herein is unclear. Here we tested a hypothetical model of reentrant feedback. We used a previous developed 2-layered feedforwardspiking network that is able to segregate figure from ground and included feedback connections. Our computer model data show that without feedback, neurons respond with regular low-frequency (∼9 Hz) bursting to a figure-ground stimulus. After including feedback the firing pattern changed into a regular (tonic) spiking pattern. In this state, we found an extra enhancement of figure responses and a further suppression of background responses resulting in a stronger figure-ground signal. Such push-pull effect was confirmed by comparing the figure-ground responses withthe responses to a homogenous texture. We propose that feedback controlsfigure-ground segregation by influencing the neural firing patterns of feedforward projecting neurons.

Feed-forward segmentation of figure-ground and assignment of border-ownership

Figure-ground is the segmentation of visual information into objects and their surrounding backgrounds. Two main processes herein are boundary assignment and surface segregation, which rely on the integration of global scene information. Recurrent processing either by intrinsic horizontal connections that connect surrounding neurons or by feedback projections from higher visual areas provide such information, and are considered to be the neural substrate for figure-ground segmentation. On the contrary, a role of feedforward projections in figure-ground segmentation is unknown. To have a better understanding of a role of feedforward connections in figure-ground organization, we constructed a feedforward spiking model using a biologically plausible neuron model. By means of surround inhibition our simple 3-layered model performs figure-ground segmentation and one-sided border-ownership coding. We propose that the visual system uses feed forward suppression for figure-ground segmentation and border-ownership assignment.

The universal fuzzy logical framework of neural circuits and its application in modeling primary visual cortex

Analytical study of large-scale nonlinear neural circuits is a difficult task. Here we analyze the function of neural systems by probing the fuzzy logical framework of the neural cells' dynamical equations. Although there is a close relation between the theories of fuzzy logical systems and neural systems and many papers investigate this subject, most investigations focus on finding new functions of neural systems by hybridizing fuzzy logical and neural system. In this paper, the fuzzy logical framework of neural cells is used to understand the nonlinear dynamic attributes of a common neural system by abstracting the fuzzy logical framework of a neural cell. Our analysis enables the educated design of network models for classes of computation. As an example, a recurrent network model of the primary visual cortex has been built and tested using this approach.

Unsupervised segmentation with dynamical units

In this paper, we present a novel network to separate mixtures of inputs that have been previously learned. A significant capability of the network is that it segments the components of each input object that most contribute to its classification. The network consists of amplitude-phase units that can synchronize their dynamics, so that separation is determined by the amplitude of units in an output layer, and segmentation by phase similarity between input and output layer units. Learning is unsupervised and based on a Hebbian update, and the architecture is very simple. Moreover, efficient segmentation can be achieved even when there is considerable superposition of the inputs. The network dynamics are derived from an objective function that rewards sparse coding in the generalized amplitude-phase variables. We argue that this objective function can provide a possible formal interpretation of the binding problem and that the implementation of the network architecture and dynamics is biologically plausible.