Mona Mathur

A Cooperation and Competition Based Simple Cell Receptive Field Model and Study of Feed-Forward Linear and Nonlinear Contributions to Orientation Selectivity

We present a model for development of orientation selectivity in layer IV simple cells. Receptive field (RF) development in the model, is determined by diffusive cooperation and resource limited competition guided axonal growth and retraction in geniculocortical pathway. The simulated cortical RFs resemble experimental RFs. The receptive field model is incorporated in a three-layer visual pathway model consisting of retina, LGN and cortex. We have studied the effect of activity dependent synaptic scaling on orientation tuning of cortical cells. The mean value of hwhh (half width at half the height of maximum response) in simulated cortical cells is 58° when we consider only the linear excitatory contribution from LGN. We observe a mean improvement of 22.8° in tuning response due to the non-linear spiking mechanisms that include effects of threshold voltage and synaptic scaling factor.

Parametric Video Compression Scheme Using AR Based Texture Synthesis

In this paper a video coding scheme based on parametric compression of texture is proposed. Each macro block is characterized either as an edge block, or as a non edge block containing texture. The non edge blocks are coded by modeling them as an auto-regressive process (AR). By applying the AR model in spatio-temporal domain, we ensure both spatial as well as temporal consistency. Edge blocks are encoded using the standard H.264/AVC. The proposed algorithm achieves upto 54.52% more compression as compared to the standard H.264/AVC at similar visual quality.

Study of spatial frequency selectivity and its spatial organization in the visual cortex through a feedforward model

A purely feedforward model has been shown to produce realistic simple cell receptive fields (RFs) that show smooth transitions between subregions and fade off gradually at the boundaries (J. Comp. Neurosci. 14 (2003) 211). Here, we show that the modeled cells also capture a wide range of spatial frequency properties of cortical cells. The shape, size and number of the subregions in the RF is found to be an important parameter in determining the frequency selectivity of the cell. The spatial frequency maps obtained through the model show a continuous distribution of spatial frequency preference across the modeled cortex and pinwheels largely co-localize with the extremes of the spatial frequency domains.

Restricted affine motion compensation in video coding using particle filtering

We propose a novel particle filter-based motion compensation strategy for video coding. We use a higher order linear model in place of the traditional translational model used in standards such as H.264. The measurement/observation process in the particle filter is a computationally efficient mechanism as opposed to traditional search methods. We use a multi-resolution framework for efficient parameter estimation. Results of our experimentation show reduced residual energy and better PSNR as compared to traditional video coding methods, especially in regions of complex motion such as zooming and rotation.

A Model That Captures Receptive Field Properties of Orientation Selective Neurons in the Visual Cortex

A purely feedforward model has been shown to produce realistic simple cell receptive fields (RFs). The modeled cells capture a wide range of receptive field properties of orientation selective cortical cells in the primary visual cortex. We have analyzed the responses of 72 nearby cell pairs to study which RF properties are clustered. Orientation preference shows strongest clustering and RF phase the least clustering. Our results agree well with experimental data (DeAngelis et al, 1999, Swindale et al, 2003).

Restricted affine motion compensation and estimation in video coding with particle filtering and importance sampling: a multi-resolution approach

In this paper, we propose a multi-resolution affine block-based tracker for motion estimation and compensation, compatible with existing video coding standards such as H.264 and HEVC. We propose three modifications to traditional motion compensation techniques in video coding standards such as H.264 and HEVC. First, we replace traditional search methods with an efficient particle filtering-based method, which incorporates information from both spatial and temporal continuity. Second, we use a higher order linear model in place of the traditional translation motion model in these standards to efficiently represent complex motions such as rotation and zoom. Third, we propose a multi-resolution framework that enables efficient parameter estimation. Results of extensive experimentation show reduced residual energy and better Peak Signal-to-Noise Ratio (PSNR, hereafter) as compared to H.264/HEVC for instance, especially in regions of complex motion such as zooming and rotation.

RBF based spatio-temporal representation technique for video compression

Parametric coding is a technique in which data is processed to extract meaningful information and then representing it compactly using appropriate parameters. Parametric Coding exploits redundancy in information to provide a very compact representation and thus achieves very high compression ratios. However, this is achieved at the cost of higher computation complexity. This disadvantage is now being offset by the availability of high speed processors, thus making it possible to exploit the high compression ratios of the parametric video coding techniques. In this paper a novel idea for efficient parametric representation of video is proposed. We perform Oct-Tree Decomposition on a video stack, followed by parameter extraction using Radial Basis Function Networks (RBFN) to achieve exceptionally high compression ratios, even higher than the state of art H.264 codec. The proposed technique exploits spatial-temporal redundancy and therefore inherently achieves multiframe prediction.

Orientation Map Emerges in Parallel with the Formation of Receptive Fields in a Feedforward Neurotrophic Model

A feed-forward neurotrophic model has been shown to generate realistic receptive field (RF) profiles for simple cells that show smooth transitions between subregions and fade off gradually at the boundaries [1]. RF development in the neurotrophic model is determined by diffusive cooperation and resource limited competition guided axonal growth and retraction in the geniculocortical pathway. Simple cells developed through the model are selective for orientation (OR) [1] and capture a wide range of spatial frequency properties of cortical cells [2]. Here, we show that the development of spatial receptive structure of the cells through the phenomena of competition and cooperation is also accompanied with formation of an orientation map (ORmap). Once these maps appear they remain stable.

Receptive Fields and Orientation Selectivity of SimpleCells in Visual Cortex

We present a three-layer feedforward visual pathway model consisting of retina, LGN and cortex for development of orientation selectivity in layer 4 simple cells. Receptive field (RF) development in the model, is determined by diffusive cooperation and resource limited competition guided axonal growth and retraction in geniculocortical pathway. The simulated cortical RFs resemble experimental RFs. In the feed forward model when only the linear contribution from LGN was considered we obtained a mean hwhh (half width at half the maximum response) of 58°. The mean hwhh improved to 37° when nonlinearity in spiking mechanism was incorporated. The hwhh in the cells improved by another 9° when a push-pull effect was included. The tuning in the simulated cells matches reported tuning in layer 4 simple cells.