A robust event-driven approach to always-on object recognition

<div> <div> <div> <p>We propose a neuromimetic architecture able to perform always-on pattern recognition. To achieve this, we extended an existing event-based algorithm [1], which introduced novel spatio-temporal features as a Hierarchy Of Time-Surfaces (HOTS). Built from asynchronous events acquired by a neuromorphic camera, these time surfaces allow to code the local dynamics of a visual scene and to create an efficient event-based pattern recognition architecture. Inspired by neuroscience, we extended this method to increase its performance. Our first contribution was to add a homeostatic gain control on the activity of neurons to improve the learning of spatio-temporal patterns [2]. A second contribution is to draw an analogy between the HOTS algorithm and Spiking Neural Networks (SNN). Following that analogy, our last contribution is to modify the classification layer and remodel the offline pattern categorization method previously used into an online and event-driven one. This classifier uses the spiking output of the network to define novel time surfaces and we then perform online classification with a neuromimetic implementation of a multinomial logistic regression. Not only do these improvements increase consistently the performances of the network, they also make this event-driven pattern recognition algorithm online and bio-realistic. Results were validated on different datasets: DVS barrel [3], Poker-DVS [4] and N-MNIST [5]. We foresee to develop the SNN version of the method and to extend this fully event-driven approach to more naturalistic tasks, notably for always-on, ultra-fast object categorization. </p> </div> </div> </div>

A robust event-driven approach to always-on object recognition

<div> <div> <div> <p>We propose a neuromimetic architecture able to perform always-on pattern recognition. To achieve this, we extended an existing event-based algorithm [1], which introduced novel spatio-temporal features as a Hierarchy Of Time-Surfaces (HOTS). Built from asynchronous events acquired by a neuromorphic camera, these time surfaces allow to code the local dynamics of a visual scene and to create an efficient event-based pattern recognition architecture. Inspired by neuroscience, we extended this method to increase its performance. Our first contribution was to add a homeostatic gain control on the activity of neurons to improve the learning of spatio-temporal patterns [2]. A second contribution is to draw an analogy between the HOTS algorithm and Spiking Neural Networks (SNN). Following that analogy, our last contribution is to modify the classification layer and remodel the offline pattern categorization method previously used into an online and event-driven one. This classifier uses the spiking output of the network to define novel time surfaces and we then perform online classification with a neuromimetic implementation of a multinomial logistic regression. Not only do these improvements increase consistently the performances of the network, they also make this event-driven pattern recognition algorithm online and bio-realistic. Results were validated on different datasets: DVS barrel [3], Poker-DVS [4] and N-MNIST [5]. We foresee to develop the SNN version of the method and to extend this fully event-driven approach to more naturalistic tasks, notably for always-on, ultra-fast object categorization. </p> </div> </div> </div>

Anti-Roll Characteristics of Marine Gyrostabilizer Based on Adaptive Control and Hydrodynamic Simulation

The gyrostabilizer produces the anti-roll effect through the precession output moment generated by a high-speed rotating flywheel. As a floating-base multi-body system composed of ship and gyrostabilizer, the recent research that has only focused on the control strategies or multi-body dynamics is obviously not comprehensive. This study presents an adaptive controller based on the variable gain control strategy for a marine gyrostabilizer installed on a port salvage tug. The variable gain control strategy controlled the flywheel precession output moment of the gyrostabilizer and thereby of the precession process, to reduce the ship roll motion effectively. Furthermore, a full-system hydrodynamic model of a gyrostabilizer-ship-wave based on three-dimensional numerical wave flume technology was innovatively established to evaluate its anti-roll performance under irregular wave conditions. The simulation results show that, for the sea state considered, the increase of spin rate of gyrostabilizer flywheel improved the anti-roll effect significantly. The average anti-roll rate of the gyrostabilizer decreased with the increase of significant wave height, wave period and wave encounter angle.

The Experiences of People With Advanced Cancer and Professionals Participating in a Program With Focus on Rehabilitation and Palliative Care

The study’s aim is to explore the experiences of people with advanced cancer and professionals participating in a program with focus on rehabilitation and palliative care. The study is based on two adjusted rehabilitation programs for 33 people with advanced cancer and 12 professionals. An observational study was conducted among the participants and two focus group interviews with 10 professionals. The analytic themes were “Lightness and happiness to gain control in everyday life,” “Community and closeness,” and “Training as a happiness and changing agent.” The activities had to support physical functions and everyday activities promoting body identity and well-being as well as emotions like closeness, lightness, and happiness in groups with like-minded people and at home with a partner and other family members. These activities and theory of emotions and body can expand the understanding of palliative care and rehabilitation as separated or integrated perspectives theoretical and in practice.

Disturbed Balance of Inhibitory Signaling Links Hearing Loss and Cognition

Neuronal hyperexcitability in the central auditory pathway linked to reduced inhibitory activity is associated with numerous forms of hearing loss, including noise damage, age-dependent hearing loss, and deafness, as well as tinnitus or auditory processing deficits in autism spectrum disorder (ASD). In most cases, the reduced central inhibitory activity and the accompanying hyperexcitability are interpreted as an active compensatory response to the absence of synaptic activity, linked to increased central neural gain control (increased output activity relative to reduced input). We here suggest that hyperexcitability also could be related to an immaturity or impairment of tonic inhibitory strength that typically develops in an activity-dependent process in the ascending auditory pathway with auditory experience. In these cases, high-SR auditory nerve fibers, which are critical for the shortest latencies and lowest sound thresholds, may have either not matured (possibly in congenital deafness or autism) or are dysfunctional (possibly after sudden, stressful auditory trauma or age-dependent hearing loss linked with cognitive decline). Fast auditory processing deficits can occur despite maintained basal hearing. In that case, tonic inhibitory strength is reduced in ascending auditory nuclei, and fast inhibitory parvalbumin positive interneuron (PV-IN) dendrites are diminished in auditory and frontal brain regions. This leads to deficits in central neural gain control linked to hippocampal LTP/LTD deficiencies, cognitive deficits, and unbalanced extra-hypothalamic stress control. Under these conditions, a diminished inhibitory strength may weaken local neuronal coupling to homeostatic vascular responses required for the metabolic support of auditory adjustment processes. We emphasize the need to distinguish these two states of excitatory/inhibitory imbalance in hearing disorders: (i) Under conditions of preserved fast auditory processing and sustained tonic inhibitory strength, an excitatory/inhibitory imbalance following auditory deprivation can maintain precise hearing through a memory linked, transient disinhibition that leads to enhanced spiking fidelity (central neural gain⇑) (ii) Under conditions of critically diminished fast auditory processing and reduced tonic inhibitory strength, hyperexcitability can be part of an increased synchronization over a broader frequency range, linked to reduced spiking reliability (central neural gain⇓). This latter stage mutually reinforces diminished metabolic support for auditory adjustment processes, increasing the risks for canonical dementia syndromes.

Teleoperated Surgical Robot Security

Technology has greatly increased the availability of medical procedures in remote locations that are difficult to access, such as battlefields. Teleoperated surgical robots can be used to perform surgeries on patients over the internet in remote locations. A surgeon can remotely operate the robot to perform a procedure in another room or in a different continent. However, security technology has not yet caught up to these cyber-physical devices. There exist potential cybersecurity attacks on these medical devices that could expose a patient to danger in contrast to traditional surgery. Hence, the security of the system is very important. A malicious actor can gain control of the device and potentially threaten the life of a patient. In this chapter, the authors conduct a survey of potential attack vectors a malicious actor could exploit to deny service to the device, gain control of the device, and steal patient data. Furthermore, after the vulnerability analysis, the authors provide mitigation techniques to limit the risk of these attack vectors.

A large gain variable range, high linearity, lownoise, low DC offset VGAs used in BD system

In this paper, a large gain variable range, high linearity, low noise, low DC offset VGAs with a simple gain-dB variable circuit are introduced. In the VGAs chain, the last and the first VGAs employ Bipolar transistors, to improve the linearity and noise characteristics. And the middle three stages VGAs employ MOS transistors. The whole circuitry is designed in 0.35um BiCMOS process, including variable gain amplifiers (VGAs) , fixed gain amplifiers , gain control and DC offset cancellation parts. The automatic gain control loop (AGC) provides a process independent gain variable range of 60dB (including 50dB gain-dB-linearity variable range), with a 200us loop lock time, the VGAs provide a 73dB largest gain, the THD is less than 1% at a 1V(P-P) output level; the equivalent output integral noise is 0.011v/√hz@20MHz bandwidth. The whole area is 1173um*494 um, and the power is 7.1mA at 3.3V signal supply voltage.

Eliminating Group Agency

Aggregating individuals’ consistent attitudes might produce inconsistent collective attitudes. Some groups therefore need the capacity to form attitudes that are irreducible to those of their members. Such groups, group-agent realists argue, are agents in control of their own attitude formation. In this paper, however, I show how group-agent realism overlooks the important fact that groups consist of strategically interacting agents. Only by eliminating group agency from our social explanations can we see how individuals vote strategically to gain control of their groups and produce collective attitudes we cannot make sense of if we treat groups as agents.