In this paper, we present a novel scale- and rotation-invariant interest point detector and descriptor, coined SURF (Speeded Up Robust Features). It approximates or even outperforms previously proposed schemes with respect to repeatability, distinctiveness, and robustness, yet can be computed and compared much faster.

This is achieved by relying on integral images for image convolutions; by building on the strengths of the leading existing detectors and descriptors (in casu, using a Hessian matrix-based measure for the detector, and a distribution-based descriptor); and by simplifying these methods to the essential. This leads to a combination of novel detection, description, and matching steps. The paper presents experimental results on a standard evaluation set, as well as on imagery obtained in the context of a real-life object recognition application. Both show SURF's strong performance.

/pdf/surf-speeded-up-robust-features-1fje09jteo.pdf

SURF: speeded up robust features

The study of decision making spans such varied fields as neuroscience, psychology, economics, statistics, political science, and computer science. Despite this diversity of applications, most decisions share common elements including deliberation and commitment. Here we evaluate recent progress in understanding how these basic elements of decision formation are implemented in the brain. We focus on simple decisions that can be studied in the laboratory but emphasize general principles likely to extend to other settings.

/pdf/the-neural-basis-of-decision-making-2p5agsty56.pdf

The Neural Basis of Decision Making

Neurons in the brain communicate by short electrical pulses, the so-called action potentials or spikes. How can we understand the process of spike generation? How can we understand information transmission by neurons? What happens if thousands of neurons are coupled together in a seemingly random network? How does the network connectivity determine the activity patterns? And, vice versa, how does the spike activity influence the connectivity pattern? These questions are addressed in this 2002 introduction to spiking neurons aimed at those taking courses in computational neuroscience, theoretical biology, biophysics, or neural networks. The approach will suit students of physics, mathematics, or computer science; it will also be useful for biologists who are interested in mathematical modelling. The text is enhanced by many worked examples and illustrations. There are no mathematical prerequisites beyond what the audience would meet as undergraduates: more advanced techniques are introduced in an elementary, concrete fashion when needed.

/pdf/spiking-neuron-models-single-neurons-populations-plasticity-58gzrd6ih6.pdf

Spiking Neuron Models: Single Neurons, Populations, Plasticity

The linear transform model of functional magnetic resonance imaging (fMRI) hypothesizes that fMRI responses are proportional to local average neural activity averaged over a period of time. This work reports results from three empirical tests that support this hypothesis. First, fMRI responses in human primary visual cortex (V1) depend separably on stimulus timing and stimulus contrast. Second, responses to long-duration stimuli can be predicted from responses to shorter duration stimuli. Third, the noise in the fMRI data is independent of stimulus contrast and temporal period. Although these tests can not prove the correctness of the linear transform model, they might have been used to reject the model. Because the linear transform model is consistent with our data, we proceeded to estimate the temporal fMRI impulse-response function and the underlying (presumably neural) contrast-response function of human V1.

/pdf/linear-systems-analysis-of-functional-magnetic-resonance-4992qtj053.pdf

Linear Systems Analysis of Functional Magnetic Resonance Imaging in Human V1

Neuronal Synchrony: A Versatile Code for the Definition of Relations?

The possibility has been explored that a subset of physiologically identifiable cells in the visual cortex is especially suited for the processing of stereoscopic depth information. First, characteristics of a disparity detector that would be useful for such processing were outlined. Then, a method was devised by which detailed binocular response data were obtained from cortical cells. In addition, a model of the disparity detector was developed that includes a plausible hierarchical arrangement of cortical cells. Data from the cells compare well with the requirements for the archetypal disparity detector and are in excellent agreement with the predictions of the model. These results demonstrate that a specific type of cortical neuron exhibits the desired characteristics of a disparity detector.

https://science.sciencemag.org/content/sci/249/4972/1037.full.pdf

Stereoscopic depth discrimination in the visual cortex: neurons ideally suited as disparity detectors

We have examined the idea that the adaptation of cortical neurons to local contrast levels in a visual stimulus is functionally advantageous. Specifically, cortical cells may have large differential contrast sensitivity as a result of adjustments that center a limited response range around a mean level of contrast. To evaluate this notion, we measured contrast-response functions of cells in striate cortex while systematically adapting them to different contrast levels of stimulus gratings. For the majority of cortical neurons tested, the results of this basic experiment show that contrast-response functions shift laterally along a log-contrast axis so that response functions match mean contrast levels in the stimulus. This implies a contrast-dependent change in the gain of the cell's contrast-response relationship. We define this process as contrast gain control. The degree to which this contrast adjustment occurs varies considerably from cell to cell. There are no obvious differences regarding cell type (simple vs. complex) or laminar distribution. Contrast gain control is almost certainly a cortical function, since lateral geniculate cells and fibers exhibit only minimal effects. Tests presented in the accompanying paper (37) provide additional evidence on the cortical origin of the process. In another series of experiments, the effect of contrast adaptation on physiological estimates of contrast sensitivity was evaluated. Sustained adaptation to contrast levels as low as 3% was capable of nearly doubling the thresholds of most of the cells tested. Adaptation may therefore be an important factor in determinations of the contrast sensitivity of cortical neurons. We tested the spatial extent of the mechanisms responsible for these gain-control effects by attempting to adapt cells using both a large grating and a grating patch limited to that portion of a cell's receptive field from which excitatory discharges could be elicited directly (the central discharge region). Adaptation was found to be an exclusive property of the central region. This held even in the case of hypercomplex cells, which received strong influences from surrounding regions of the visual field. Finally, we measured the time course of contrast adaptation. We found the process to be rather slow, with a mean time constant of approximately 6 s. Once again, there was considerable variability in this value from cell to cell.

Contrast gain control in the cat's visual system

1. The classically defined receptive field of a visual neuron is the area of visual space over which the cell responds to visual stimuli. It is well established, however, that the discharge produced by an optimal stimulus can be modulated by the presence of additional stimuli that by themselves do not produce any response. This study examines inhibitory influences that originate from areas located outside of the classical (i.e., excitatory) receptive field. Previous work has shown that for some cells the response to a properly oriented bar of light becomes attenuated when the bar extends beyond the receptive field, a phenomenon known as end-inhibition (or length tuning). Analogously, it has been shown that increasing the number of cycles of a drifting grating stimulus may also inhibit the firing of some cells, an effect known as side-inhibition (or width tuning). Very little information is available, however, about the relationship between end- and side-inhibition. We have examined the spatial organization and tuning characteristics of these inhibitory effects by recording extracellularly from single neurons in the cat's striate cortex (Area 17). 2. For each cortical neuron, length and width tuning curves were obtained with the use of rectangular patches of drifting sinusoidal gratings that have variable length and width. Results from 82 cells show that the strengths of end- and side-inhibition tend to be correlated. Most cells that exhibit clear end-inhibition also show a similar degree of side-inhibition. For these cells, the excitatory receptive field is surrounded on all sides by inhibitory zones. Some cells exhibit only end- or side-inhibition, but not both. Data for 28 binocular cells show that length and width tuning curves for the dominant and nondominant eyes tend to be closely matched. 3. We also measured tuning characteristics of end- and side-inhibition. To obtain these data, the excitatory receptive field was stimulated with a grating patch having optimal orientation, spatial frequency, and size, whereas the end- or side-inhibitory regions were stimulated with patches of gratings that had a variable parameter (such as orientation). Results show that end- and side-inhibition tend to be strongest at the orientation and spatial frequency that yield maximal excitation. However, orientation and spatial frequency tuning curves for inhibition are considerably broader than those for excitation, suggesting that inhibition is mediated by a pool of neurons.(ABSTRACT TRUNCATED AT 400 WORDS)

Length and width tuning of neurons in the cat's primary visual cortex

https://www.cell.com/trends/neurosciences/pdf/0166-2236(95)94496-R.pdf

Receptive-field dynamics in the central visual pathways

1. Most studies of cortical neurons have focused on the spatial structure of receptive fields. For a more complete functional description of these neurons, it is necessary to consider receptive-field structure in the joint domain of space and time. We have studied the spatiotemporal receptive-field structure of 233 simple cells recorded from the striate cortex of adult cats and kittens at 4 and 8 wk postnatal. The dual goal of this study is to provide a detailed quantitative description of spatiotemporal receptive-field structure and to compare the developmental time courses of spatial and temporal response properties. 2. Spatiotemporal receptive-field profiles have been measured with the use of a reverse correlation method, in which we compute the cross-correlation between a neuron's response and a random sequence of small, briefly presented bright and dark stimuli. The receptive-field profiles of some simple cells are space-time separable, meaning that spatial and temporal response characteristics can be dissociated. Other cells have receptive-field profiles that are space-time inseparable. In these cases, a particular spatial location cannot be designated, unambiguously, as belonging to either an on or off subregion. However, separate on and off subregions may be clearly distinguished in the joint space-time domain. These subregions are generally tilted along an oblique axis. 3. Our observations show that spatial and temporal aspects of receptive-field structure mature with clearly different time courses. By 4 wk postnatal, the spatial symmetry and periodicity of simple-cell receptive fields have reached maturity. The spatial extent (or size) of these receptive fields is adult-like by 8 wk postnatal. In contrast, the response latency and time duration of spatiotemporal receptive fields do not mature until well beyond 8 wk postnatal. 4. By applying Fourier analysis to spatiotemporal receptive-field profiles, we have examined the postnatal development of spatial and temporal selectivity in the frequency domain. By 8 wk postnatal, spatial frequency tuning has clearly reached maturity. On the contrary, temporal frequency selectivity remains markedly immature at 8 wk. We have also examined the joint distribution of optimal spatial and temporal frequencies. From 4 wk postnatal until 8 wk postnatal, the range of optimal spatial frequencies increases substantially, whereas the range of optimal temporal frequencies remains largely unchanged. From 8 wk postnatal until adulthood, there is a large increase in optimal temporal frequencies for cells tuned to low spatial frequencies. For cells tuned to high spatial frequencies, the distribution of optimal temporal frequencies does not change much beyond 8 wk postnatal.(ABSTRACT TRUNCATED AT 400 WORDS)

Izumi Ohzawa

Papers

Stereoscopic depth discrimination in the visual cortex: neurons ideally suited as disparity detectors

Contrast gain control in the cat's visual system

Length and width tuning of neurons in the cat's primary visual cortex

Receptive-field dynamics in the central visual pathways

Spatiotemporal organization of simple-cell receptive fields in the cat's striate cortex. I: General characteristics and postnatal development