HFirst: A Temporal Approach to Object Recognition.

This paper introduces a spiking hierarchical model for object recognition which utilizes the precise timing information inherently present in the output of biologically inspired asynchronous address event representation (AER) vision sensors. The asynchronous nature of these systems frees computation and communication from the rigid predetermined timing enforced by system clocks in conventional systems. Freedom from rigid timing constraints opens the possibility of using true timing to our advantage in computation. We show not only how timing can be used in object recognition, but also how it can in fact simplify computation. Specifically, we rely on a simple temporal-winner-take-all rather than more computationally intensive synchronous operations typically used in biologically inspired neural networks for object recognition. This approach to visual computation represents a major paradigm shift from conventional clocked systems and can find application in other sensory modalities and computational tasks. We showcase effectiveness of the approach by achieving the highest reported accuracy to date (97.5% ± 3.5%) for a previously published four class card pip recognition task and an accuracy of 84.9% ± 1.9% for a new more difficult 36 class character recognition task.

What can neuromorphic event-driven precise timing add to spike-based pattern recognition?

This letter introduces a study to precisely measure what an increase in spike timing precision can add to spike-driven pattern recognition algorithms. The concept of generating spikes from images by converting gray levels into spike timings is currently at the basis of almost every spike-based modeling of biological visual systems. The use of images naturally leads to generating incorrect artificial and redundant spike timings and, more important, also contradicts biological findings indicating that visual processing is massively parallel, asynchronous with high temporal resolution. A new concept for acquiring visual information through pixel-individual asynchronous level-crossing sampling has been proposed in a recent generation of asynchronous neuromorphic visual sensors. Unlike conventional cameras, these sensors acquire data not at fixed points in time for the entire array but at fixed amplitude changes of their input, resulting optimally sparse in space and time-pixel individually and precisely timed only if new, (previously unknown) information is available (event based). This letter uses the high temporal resolution spiking output of neuromorphic event-based visual sensors to show that lowering time precision degrades performance on several recognition tasks specifically when reaching the conventional range of machine vision acquisition frequencies (30-60 Hz). The use of information theory to characterize separability between classes for each temporal resolution shows that high temporal acquisition provides up to 70% more information that conventional spikes generated from frame-based acquisition as used in standard artificial vision, thus drastically increasing the separability between classes of objects. Experiments on real data show that the amount of information loss is correlated with temporal precision. Our information-theoretic study highlights the potentials of neuromorphic asynchronous visual sensors for both practical applications and theoretical investigations. Moreover, it suggests that representing visual information as a precise sequence of spike times as reported in the retina offers considerable advantages for neuro-inspired visual computations.

Asynchronous Event-Based Multikernel Algorithm for High-Speed Visual Features Tracking.

This paper presents a number of new methods for visual tracking using the output of an event-based asynchronous neuromorphic dynamic vision sensor. It allows the tracking of multiple visual features in real time, achieving an update rate of several hundred kilohertz on a standard desktop PC. The approach has been specially adapted to take advantage of the event-driven properties of these sensors by combining both spatial and temporal correlations of events in an asynchronous iterative framework. Various kernels, such as Gaussian, Gabor, combinations of Gabor functions, and arbitrary user-defined kernels, are used to track features from incoming events. The trackers described in this paper are capable of handling variations in position, scale, and orientation through the use of multiple pools of trackers. This approach avoids the N(2) operations per event associated with conventional kernel-based convolution operations with N × N kernels. The tracking performance was evaluated experimentally for each type of kernel in order to demonstrate the robustness of the proposed solution.

Cognitive task fulfilment may decrease gaze control performances.

Human cognitive processing limits can lead to difficulties in performing two tasks simultaneously. In this respect, the mobilization of attentional resources seems to be more important in voluntary than in reflexive visuo-oculomotor movements. With this in mind, this study aimed to determine the differentiated effects of carrying out cognitive tasks on reflexive and voluntary movements generated by the visuo-oculomotor system. Eye movements were recorded with a videonystagmography system in 20 healthy adults. Prosaccade and antisaccade latency and accuracy were determined, as was the gain of regular smooth pursuit. The cognitive tasks comprised a backward counting task in steps of seven or thirteen and were carried out during the three oculomotor tasks. Whatever the backward counting, the cognitive tasks caused an increase in latency and a decrease in accuracy in the prosaccades and antisaccades as well as a drop in smooth pursuit gain. Despite a high degree of prosaccade, antisaccade and pursuit predictability, the oculomotor parameters were altered during dual-task fulfilment, and voluntary movements were more vulnerable than reflexive movements. Cognitive task achievement during oculomotor tasks requires greater attentional resource mobilization, and the reduction in visuo-oculomotor performance is even greater since the oculomotor task places more demand on the cognitive resources. Attention is thus a crucial cognitive process for the maintenance of adequate reflexive and voluntary oculomotor performance. Concurrently performing tasks have broader implications in motor coordination degradation understanding in situations like handheld phone when driving or in management of organisational and environmental constraints in occupational activities.

Polysaccharide-induced order-to-order transitions in lyotropic liquid crystals.

We report on the order-to-order transitions of lyotropic liquid crystals formed by self-assembled monogylcerides and water in the presence of polysaccharides of various molecular weights. The phase diagram of monoglyceride-water-polysaccharide systems, their morphology, and the topology of liquid crystalline structures were determined by combining optical cross-polarization, oscillatory shear rheometry, and small-angle X-ray scattering. The presence of hydrophilic mono-, oligo-, and polysaccharides in the water domains of liquid crystalline phases resulted in a general decrease of the cubic-to-hexagonal transition temperature. Provided that the sugar could fit within the water channels, the decrease was observed to be dependent on the polysaccharide concentration but independent of its molecular weight. For isotropic bicontinuous cubic phases, monomeric sugars such as glucose were reported to shrink the lattice parameter of the structure without inducing phase transitions. However, when a polymeric form of glucose was used, such as dextran, transitions from the gyroidal Ia3d cubic phase to double diamond Pn3m cubic phases were observed at well-defined molecular weights of polysaccharide. These results were interpreted in terms of size exclusions of polymer sugars by the water domains of the liquid crystal phases as well as the different topologies of water channels. Molecular dynamics simulations of polysaccharides in the water environment were performed to support these findings.

Shear rheology of lyotropic liquid crystals: a case study.

We have investigated the rheological properties of lyotropic liquid crystals (LCs) formed by self-assembled neutral lipids and water, their relationship with the topology of the structure, and their dependence on temperature and water content. The phase diagram of a representative monoglyceride-water system, determined by combining cross-polarized optical microscopy and small-angle X-ray scattering (SAXS), included four structures: lamellar, hexagonal, gyroid bicontinuous cubic (Ia3d), and double diamond bicontinuous cubic (Pn3m), as well as several regions of two-phase coexistence of some of the above structures. Rheology in the linear viscoelastic regime revealed a specific signature that was characteristic of the topology of each structure considered. The order-order transitions lamellar-to-cubic and cubic-to-hexagonal, as well as the order-disorder transitions from each LC to an isotropic fluid, were easily identified by following the development of the storage and loss moduli, G' and G'', respectively. The viscoelastic properties of both bicontinuous cubic phases were shown to be strongly frequency-dependent, following a pseudo-Maxwell behavior, with multiple relaxation times. Cubic-to-cubic transitions were nicely captured by scaling the longest relaxation time, tau, with either temperature or water volume fraction. Therefore, the set of the three main parameters used to establish the rheological behavior of the structure, that is, G', G'', and relaxation time, tau, constitutes a consistent ensemble to identify the structures of the liquid crystal. Finally, relaxation spectra, extracted for all liquid crystalline phases, allowed an additional possible identification criterion of the various structures considered.