Showing 1–14 of 14 results for author: Meijer, R

Hardware-aware training of models with synaptic delays for digital event-driven neuromorphic processors

Authors: Alberto Patino-Saucedo, Roy Meijer, Amirreza Yousefzadeh, Manil-Dev Gomony, Federico Corradi, Paul Detteter, Laura Garrido-Regife, Bernabe Linares-Barranco, Manolis Sifalakis

Abstract: Configurable synaptic delays are a basic feature in many neuromorphic neural network hardware accelerators. However, they have been rarely used in model implementations, despite their promising impact on performance and efficiency in tasks that exhibit complex (temporal) dynamics, as it has been unclear how to optimize them. In this work, we propose a framework to train and deploy, in digital neur… ▽ More Configurable synaptic delays are a basic feature in many neuromorphic neural network hardware accelerators. However, they have been rarely used in model implementations, despite their promising impact on performance and efficiency in tasks that exhibit complex (temporal) dynamics, as it has been unclear how to optimize them. In this work, we propose a framework to train and deploy, in digital neuromorphic hardware, highly performing spiking neural network models (SNNs) where apart from the synaptic weights, the per-synapse delays are also co-optimized. Leveraging spike-based back-propagation-through-time, the training accounts for both platform constraints, such as synaptic weight precision and the total number of parameters per core, as a function of the network size. In addition, a delay pruning technique is used to reduce memory footprint with a low cost in performance. We evaluate trained models in two neuromorphic digital hardware platforms: Intel Loihi and Imec Seneca. Loihi offers synaptic delay support using the so-called Ring-Buffer hardware structure. Seneca does not provide native hardware support for synaptic delays. A second contribution of this paper is therefore a novel area- and memory-efficient hardware structure for acceleration of synaptic delays, which we have integrated in Seneca. The evaluated benchmark involves several models for solving the SHD (Spiking Heidelberg Digits) classification task, where minimal accuracy degradation during the transition from software to hardware is demonstrated. To our knowledge, this is the first work showcasing how to train and deploy hardware-aware models parameterized with synaptic delays, on multicore neuromorphic hardware accelerators. △ Less

Submitted 16 April, 2024; originally announced April 2024.

OpenContrails: Benchmarking Contrail Detection on GOES-16 ABI

Authors: Joe Yue-Hei Ng, Kevin McCloskey, Jian Cui, Vincent R. Meijer, Erica Brand, Aaron Sarna, Nita Goyal, Christopher Van Arsdale, Scott Geraedts

Abstract: Contrails (condensation trails) are line-shaped ice clouds caused by aircraft and are likely the largest contributor of aviation-induced climate change. Contrail avoidance is potentially an inexpensive way to significantly reduce the climate impact of aviation. An automated contrail detection system is an essential tool to develop and evaluate contrail avoidance systems. In this paper, we present… ▽ More Contrails (condensation trails) are line-shaped ice clouds caused by aircraft and are likely the largest contributor of aviation-induced climate change. Contrail avoidance is potentially an inexpensive way to significantly reduce the climate impact of aviation. An automated contrail detection system is an essential tool to develop and evaluate contrail avoidance systems. In this paper, we present a human-labeled dataset named OpenContrails to train and evaluate contrail detection models based on GOES-16 Advanced Baseline Imager (ABI) data. We propose and evaluate a contrail detection model that incorporates temporal context for improved detection accuracy. The human labeled dataset and the contrail detection outputs are publicly available on Google Cloud Storage at gs://goes_contrails_dataset. △ Less

Submitted 20 April, 2023; v1 submitted 4 April, 2023; originally announced April 2023.

Adaptive Cluster Thresholding with Spatial Activation Guarantees Using All-resolutions Inference

Authors: Xu Chen, Jelle J. Goeman, Thijmen J. P. Krebs, Rosa J. Meijer, Wouter D. Weeda

Abstract: Classical cluster inference is hampered by the spatial specificity paradox. Given the null-hypothesis of no active voxels, the alternative hypothesis states that there is at least one active voxel in a cluster. Hence, the larger the cluster the less we know about where activation in the cluster is. Rosenblatt et al. (2018) proposed a post-hoc inference method, All-resolutions Inference (ARI), that… ▽ More Classical cluster inference is hampered by the spatial specificity paradox. Given the null-hypothesis of no active voxels, the alternative hypothesis states that there is at least one active voxel in a cluster. Hence, the larger the cluster the less we know about where activation in the cluster is. Rosenblatt et al. (2018) proposed a post-hoc inference method, All-resolutions Inference (ARI), that addresses this paradox by estimating the number of active voxels of any brain region. ARI allows users to choose arbitrary brain regions and returns a simultaneous lower confidence bound of the true discovery proportion (TDP) for each of them, retaining control of the family-wise error rate. ARI does not, however, guide users to regions with high enough TDP. In this paper, we propose an efficient algorithm that outputs all maximal supra-threshold clusters, for which ARI gives a TDP lower confidence bound that is at least a chosen threshold, for any number of thresholds that need not be chosen a priori nor all at once. After a preprocessing step in linearithmic time, the algorithm only takes linear time in the size of its output. We demonstrate the algorithm with an application to two fMRI datasets. For both datasets, we found several clusters whose TDP confidently meets or exceeds a given threshold in less than a second. △ Less

Submitted 10 May, 2023; v1 submitted 27 June, 2022; originally announced June 2022.

A shortcut for Hommel's procedure in linearithmic time

Authors: Rosa Meijer, Thijmen Krebs, Aldo Solari, Jelle Goeman

Abstract: Hommel's and Hochberg's procedures for familywise error control are both derived as shortcuts in a closed testing procedure with the Simes local test. Hommel's shortcut is exact but takes quadratic time in the number of hypotheses. Hochberg's shortcut takes only linearithmic time, but is conservative. In this paper we present an exact shortcut in linearithmic time, combining the strengths of both… ▽ More Hommel's and Hochberg's procedures for familywise error control are both derived as shortcuts in a closed testing procedure with the Simes local test. Hommel's shortcut is exact but takes quadratic time in the number of hypotheses. Hochberg's shortcut takes only linearithmic time, but is conservative. In this paper we present an exact shortcut in linearithmic time, combining the strengths of both procedures. The novel shortcut also applies to a robust variant of Hommel's procedure that does not require the assumption of the Simes inequality. △ Less

Submitted 23 October, 2017; originally announced October 2017.

Comments: arXiv admin note: text overlap with arXiv:1611.06739

MattockFS; Page-cache and access-control concerns in asynchronous message-based forensic frameworks on the Linux platform

Authors: Rob J Meijer

Abstract: In this dissertation the feasibility of creating a page-cache efficient storage- and messaging solution with integrity geared access control for a scalable forensic framework is researched. The Open Computer Forensics Architecture (OCFA),a lab-side scalable computer forensics framework, introduced the concept of a message passing concurrency based forensic framework. Since then, the amount of per-… ▽ More In this dissertation the feasibility of creating a page-cache efficient storage- and messaging solution with integrity geared access control for a scalable forensic framework is researched. The Open Computer Forensics Architecture (OCFA),a lab-side scalable computer forensics framework, introduced the concept of a message passing concurrency based forensic framework. Since then, the amount of per-investigation data to be processed in a lab environment has continued to grow significantly while available RAM and CPU processing power combined with prohibitive cost and limited capacity of SSD solutions have shifted processing from being largely CPU constrained to being much more IO constrained. OCFA suffered from several page-cache-miss related performance issues that have grown more significant as a result of this shift. In the light of anti-forensics and general issues related to process integrity, OCFA did not leverage the power of its message passing based design to address integrity concerns. The main purpose of this dissertation is to analyze and evaluate a number of page-cache friendly technologies that could contribute to the creation of a computer forensics lab-geared scalable message-passing-concurrency based forensic framework with a significantly reduced quantity of page-cache-miss induced spurious IO operations, taking into account integrity related issues. Provenance logs from historic investigations conducted using the Open Computer Forensics Architecture were thoroughly analyzed in this study, during which several bottlenecks and design flaws in OCFA were identified. A number of strategies were devised to address these bottlenecks in future computer forensic frameworks. Finally, the most prominently page-cache related strategies were consolidated with access-control measures into a user-space file-system and low-level API prototype. △ Less

Submitted 1 March, 2017; originally announced March 2017.

Comments: dissertation, Univ College London, June 2016

The Majorana Demonstrator calibration system

Authors: N. Abgrall, I. J. Arnquist, F. T. Avignone III, A. S. Barabash, F. E. Bertrand, M. Boswell, A. W. Bradley, V. Brudanin, M. Busch, M. Buuck, T. S. Caldwell, C. D. Christofferson, P. -H. Chu, C. Cuesta, J. A. Detwiler, C. Dunagan, Yu. Efremenko, H. Ejiri, S. R. Elliott, Z. Fu, V. M. Gehman, T. Gilliss, G. K. Giovanetti, J. Goett, M. P. Green , et al. (42 additional authors not shown)

Abstract: The MAJORANA Collaboration is searching for the neutrinoless double-beta decay of the nucleus $^{76}$Ge. The MAJORANA DEMONSTRATOR is an array of germanium detectors deployed with the aim of implementing background reduction techniques suitable for a 1-tonne $^{76}$Ge-based search. The ultra low-background conditions require regular calibrations to verify proper function of the detectors. Radioact… ▽ More The MAJORANA Collaboration is searching for the neutrinoless double-beta decay of the nucleus $^{76}$Ge. The MAJORANA DEMONSTRATOR is an array of germanium detectors deployed with the aim of implementing background reduction techniques suitable for a 1-tonne $^{76}$Ge-based search. The ultra low-background conditions require regular calibrations to verify proper function of the detectors. Radioactive line sources can be deployed around the cryostats containing the detectors for regular energy calibrations. When measuring in low-background mode, these line sources have to be stored outside the shielding so they do not contribute to the background. The deployment and the retraction of the source are designed to be controlled by the data acquisition system and do not require any direct human interaction. In this paper, we detail the design requirements and implementation of the calibration apparatus, which provides the event rates needed to define the pulse-shape cuts and energy calibration used in the final analysis as well as data that can be compared to simulations. △ Less

Submitted 6 February, 2017; originally announced February 2017.

Journal ref: NIM A 872 (2017)

Simultaneous Control of All False Discovery Proportions in Large-Scale Multiple Hypothesis Testing

Authors: Jelle Goeman, Rosa Meijer, Thijmen Krebs, Aldo Solari

Abstract: Closed testing procedures are classically used for familywise error rate (FWER) control, but they can also be used to obtain simultaneous confidence bounds for the false discovery proportion (FDP) in all subsets of the hypotheses. In this paper we investigate the special case of closed testing with Simes local tests. We construct a novel fast and exact shortcut which we use to investigate the powe… ▽ More Closed testing procedures are classically used for familywise error rate (FWER) control, but they can also be used to obtain simultaneous confidence bounds for the false discovery proportion (FDP) in all subsets of the hypotheses. In this paper we investigate the special case of closed testing with Simes local tests. We construct a novel fast and exact shortcut which we use to investigate the power of this method when the number of hypotheses goes to infinity. We show that, if a minimal amount of signal is present, the average power to detect false hypotheses at any desired FDP level does not vanish. Additionally, we show that the confidence bounds for FDP are consistent estimators for the true FDP for every non-vanishing subset. For the case of a finite number of hypotheses, we show connections between Simes-based closed testing and the procedure of Benjamini and Hochberg. △ Less

Submitted 23 October, 2017; v1 submitted 21 November, 2016; originally announced November 2016.

First observation of a reactor-status effect on the beta+ decay rate of 22Na

Authors: Robert de Meijer, Albert Zondervan, Jan Stegenga, Steph Steyn, Robbie Lindsay, Milton van Rooy, Marco Tijs, Han Limburg

Abstract: In the search for an electron antineutrino detection method with sensitivity below the 1.8 MeV threshold for the inverse beta decay reaction, beta decay counting experiments with ca. 3 kBq 22Na and 60Co sources were conducted at unit #1 (2.775 GW_th) of the Koeberg Nuclear Power Station in South Africa. The setup consisted of one NaI crystal to measure de-excitation and annihilation photons associ… ▽ More In the search for an electron antineutrino detection method with sensitivity below the 1.8 MeV threshold for the inverse beta decay reaction, beta decay counting experiments with ca. 3 kBq 22Na and 60Co sources were conducted at unit #1 (2.775 GW_th) of the Koeberg Nuclear Power Station in South Africa. The setup consisted of one NaI crystal to measure de-excitation and annihilation photons associated with beta decay. Its volume and well shape were chosen to use coincidence summing in order to differentiate between electron capture and beta+ emission in 22Na. The setup was shielded from the reactor core by 8 m of uninterrupted concrete. Background radiation, responsible for ca. 1% of the total countrate with either source, increased by merely 3% when the reactor status changed from OFF to ON. Normalized countrates of three energy regions-of-interest (TOT, MED, HI) were parameterized to jointly describe the time dependence of two instrumental effects and a reactor-status step function in a least-squares regression analysis. With the 22Na source, the fractional countrate changes in the step from reactor OFF to ON were: (delA/A)_TOT = [-3.02 +- 0.14(stat) +- 0.07(syst)] x 10^-4, (delA/A)_MED = [+1.44 +- 0.42(stat) +- 0.07(syst)] x 10^-4, and (delA/A)_HI = [-2.70 +- 0.26(stat) +- 0.04(syst)] x 10^-4. The uncertainty budget is incomplete because it does not contain the possible influence from environmental factors and the finite stability of the MCA clock-oscillator. No reactor-status dependence was observed with the 60Co source. The corresponding cross sections are [1.55 +- 0.07(stat)] x 10^-25 cm^2 for EC + beta+ decay in 22Na and [0.5 +- 1.5(stat)] x 10^-26 cm^2 for beta- decay of 60Co. The negative sign for TOT and HI activity changes in 22Na points to an antineutrino related interference effect on the beta+ decay of 22Na and rules out reactor neutron induced reactions. △ Less

Submitted 5 July, 2017; v1 submitted 5 October, 2016; originally announced October 2016.

Comments: resubmitted to Physical Review C after addressing shortcomings raised during review of the original submission; 9 figures

Upper limit on the cross section for reactor antineutrinos changing 22Na decay rates

Authors: R. J. de Meijer, S. W. Steyn

Abstract: In this paper we present results of a long-term observation of the decay of 22Na in the presence of a nuclear fission reactor. The measurements were made outside the containment wall of and underneath the Koeberg nuclear power plant near Cape Town, South Africa. Antineutrino fluxes ranged from ~5*10^11 to 1.6*10^13 cm^-2 s^-1 during this period. We show that the coincidence summing technique provi… ▽ More In this paper we present results of a long-term observation of the decay of 22Na in the presence of a nuclear fission reactor. The measurements were made outside the containment wall of and underneath the Koeberg nuclear power plant near Cape Town, South Africa. Antineutrino fluxes ranged from ~5*10^11 to 1.6*10^13 cm^-2 s^-1 during this period. We show that the coincidence summing technique provides a sensitive tool to measure a change in the total decay constant as well as the branching ratio between EC and beta+ decay of 22Na to the first excited state in 22Ne. We observe a relative change in count rate between reactor-ON and reactor-OFF equal to (-0.51+/-0.11)*10^-4. After evaluating possible systematic uncertainties we conclude that the effect is either due to a hidden instrumental cause or due to an interaction between antineutrinos and the 22Na nucleus. An upper limit of ~0.03 barn has been deduced for observing any change in the decay rate of 22Na due to antineutrino interactions. △ Less

Submitted 23 September, 2014; originally announced September 2014.

Comments: 22 pages, 12 figures, 4 tables. Submitted to Astroparticle Physics on 19 August 2014

Forming the Moon from terrestrial silicate-rich material

Authors: R. J. de Meijer, V. F. Anisichkin, W. van Westrenen

Abstract: Recent high-precision measurements of the isotopic composition of lunar rocks demonstrate that the bulk silicate Earth and the Moon show an unexpectedly high degree of similarity. This is inconsistent with one of the primary results of classic dynamical simulations of the widely accepted giant impact model for the formation of the Moon, namely that most of the mass of the Moon originates from the… ▽ More Recent high-precision measurements of the isotopic composition of lunar rocks demonstrate that the bulk silicate Earth and the Moon show an unexpectedly high degree of similarity. This is inconsistent with one of the primary results of classic dynamical simulations of the widely accepted giant impact model for the formation of the Moon, namely that most of the mass of the Moon originates from the impactor, not Earth. Resolution of this discrepancy without changing the main premises of the giant impact model requires total isotopic homogenisation of Earth and impactor material after the impact for a wide range of elements including O, Si, K, Ti, Nd and W. Even if this process could explain the O isotope similarity, it is unlikely to work for the much heavier, refractory elements. Given the increasing uncertainty surrounding the giant impact model in light of these geochemical data, alternative hypotheses for lunar formation should be explored. In this paper, we revisit the hypothesis that the Moon was formed directly from terrestrial mantle material. We show that the dynamics of this scenario requires a large amount of energy, almost instantaneously generated additional energy. The only known source for this additional energy is nuclear fission. We show that it is feasible to form the Moon through the ejection of terrestrial silicate material triggered by a nuclear explosion at Earths core-mantle boundary (CMB), causing a shock wave propagating through the Earth. Hydrodynamic modelling of this scenario shows that a shock wave created by rapidly expanding plasma resulting from the explosion disrupts and expels overlying mantle and crust material. △ Less

Submitted 11 January, 2013; v1 submitted 24 January, 2010; originally announced January 2010.

Comments: 26 pages, 5 figures, 1 table

Report number: EARTH-REP 035

Assessing the feasibility and consequences of nuclear georeactors in the Earths core mantle boundary

Authors: R. J. de Meijer, W. van Westrenen

Abstract: We assess the likelihood and geochemical consequences of the presence of nuclear georeactors in the core mantle boundary region (CMB) between Earths silicate mantle and metallic core. Current geochemical models for the Earths interior predict that U and Th in the CMB are concentrated exclusively in the mineral calcium silicate perovskite (CaPv), leading to predicted concentration levels of appro… ▽ More We assess the likelihood and geochemical consequences of the presence of nuclear georeactors in the core mantle boundary region (CMB) between Earths silicate mantle and metallic core. Current geochemical models for the Earths interior predict that U and Th in the CMB are concentrated exclusively in the mineral calcium silicate perovskite (CaPv), leading to predicted concentration levels of approximately 12 ppm combined U and Th, 4.5 Ga ago if CaPv is distributed evenly throughout the CMB. Assuming a similar behaviour for primordial 244Pu provides a considerable flux of neutrons from spontaneous fission. We show that an additional concentration factor of only an order of magnitude is required to both ignite and maintain self sustaining georeactors based on fast fission. Continuously operating georeactors with a power of 5 TW can explain the observed isotopic compositions of helium and xenon in the Earths mantle. Our hypothesis requires the presence of elevated concentrations of U and Th in the CMB, and is amenable to testing by direction sensitive geoneutrino tomography. △ Less

Submitted 6 May, 2008; originally announced May 2008.

Comments: 20 pages, 10 tables, no figures

Towards Earth AntineutRino TomograpHy (EARTH)

Authors: R. J. de Meijer, F. D. Smit, F. D. Brooks, R. W. Fearick, H. J. Woertche, F. Mantovani

Abstract: The programme Earth AntineutRino TomograpHy (EARTH) proposes to build ten underground facilities each hosting a telescope. Each telescope consists of many detector modules, to map the radiogenic heat sources deep in the interior of the Earth by utilising direction sensitive geoneutrino detection. Recent hypotheses target the core-mantle boundary (CMB) as a major source of natural radionuclides a… ▽ More The programme Earth AntineutRino TomograpHy (EARTH) proposes to build ten underground facilities each hosting a telescope. Each telescope consists of many detector modules, to map the radiogenic heat sources deep in the interior of the Earth by utilising direction sensitive geoneutrino detection. Recent hypotheses target the core-mantle boundary (CMB) as a major source of natural radionuclides and therefore of radiogenic heat. A typical scale of the processes that take place at the CMB is about 200km. To observe these processes from the surface requires an angular resolution of about 3 degrees. EARTH aims at creating a high-resolution 3D-map of the radiogenic heat sources in the interior of the Earth. It will thereby contribute to a better understanding of a number of geophysical phenomena observed at the surface of the Earth. This condition requires a completely different approach from the monolithic detector systems as e.g. KamLAND. This paper presents, for such telescopes, the boundary conditions set by physics, the estimated count rates, and the first initial results from Monte Carlo simulations and laboratory experiments. The Monte Carlo simulations indicate that the large volume telescope should consist of detector modules each comprising a very large number of detector units, with a cross section of roughly a few square centimetres. The signature of an antineutrino event will be a double pulse event. One pulse arises from the slowing down of the emitted positron, the other from the neutron capture. In laboratory experiments small sized, 10B-loaded liquid scintillation detectors were investigated as candidates for direction sensitive, low-energy antineutrino detection. △ Less

Submitted 6 July, 2006; originally announced July 2006.

Comments: 13 pages, 6 figures, 1 table

Quest for the Nuclear Georeactor

Authors: R. J. de Meijer, E. R. van der Graaf, K. P. Jungmann

Abstract: The paper focuses on a proposal for an underground antineutrino antenna to further develop the dectection of these particles as a tool to map the distribution of radiogenic heat sources, such as the natural radionuclides and a hypothetical nuclear georeactor. The paper focuses on a proposal for an underground antineutrino antenna to further develop the dectection of these particles as a tool to map the distribution of radiogenic heat sources, such as the natural radionuclides and a hypothetical nuclear georeactor. △ Less

Submitted 8 April, 2004; originally announced April 2004.

Comments: accepted for Nuclear Physics News, 2004

Journal ref: Nucl.Phys.News 14 (2004) 20-25

Quest for a Nuclear Georeactor

Authors: R. J. de Meijer, E. R. van der Graaf, K. P. Jungmann

Abstract: Knowledge about the interior of our planet is mainly based on the interpretation of seismic data from earthquakes and nuclear explosions, and of composition of meteorites. Additional observations have led to a wide range of hypotheses on the heat flow from the interior to the crust, the abundance of certain noble gases in gasses vented from volcanoes and the possibility of a nuclear georeactor a… ▽ More Knowledge about the interior of our planet is mainly based on the interpretation of seismic data from earthquakes and nuclear explosions, and of composition of meteorites. Additional observations have led to a wide range of hypotheses on the heat flow from the interior to the crust, the abundance of certain noble gases in gasses vented from volcanoes and the possibility of a nuclear georeactor at the centre of the Earth. This paper focuses on a proposal for an underground laboratory to further develop antineutrinos as a tool to map the distribution of radiogenic heat sources, such as the natural radionuclides and the hypothetical nuclear georeactor. △ Less

Submitted 12 April, 2004; originally announced April 2004.

Comments: Invited talk presented at the International Symposium on Radiation Physics, Cape Town, 2003. Manuscript is submitted to Radiation Physics and Chemistry

Journal ref: Radiat.Phys.Chem. 71 (2004) 769-774