Showing 1–11 of 11 results for author: Palermo, M

Deep Inertial Pose: A deep learning approach for human pose estimation

Authors: Sara M. Cerqueira, Manuel Palermo, Cristina P. Santos

Abstract: Inertial-based Motion capture system has been attracting growing attention due to its wearability and unsconstrained use. However, accurate human joint estimation demands several complex and expertise demanding steps, which leads to expensive software such as the state-of-the-art MVN Awinda from Xsens Technologies. This work aims to study the use of Neural Networks to abstract the complex biomecha… ▽ More Inertial-based Motion capture system has been attracting growing attention due to its wearability and unsconstrained use. However, accurate human joint estimation demands several complex and expertise demanding steps, which leads to expensive software such as the state-of-the-art MVN Awinda from Xsens Technologies. This work aims to study the use of Neural Networks to abstract the complex biomechanical models and analytical mathematics required for pose estimation. Thus, it presents a comparison of different Neural Network architectures and methodologies to understand how accurately these methods can estimate human pose, using both low cost(MPU9250) and high end (Mtw Awinda) Magnetic, Angular Rate, and Gravity (MARG) sensors. The most efficient method was the Hybrid LSTM-Madgwick detached, which achieved an Quaternion Angle distance error of 7.96, using Mtw Awinda data. Also, an ablation study was conducted to study the impact of data augmentation, output representation, window size, loss function and magnetometer data on the pose estimation error. This work indicates that Neural Networks can be trained to estimate human pose, with results comparable to the state-of-the-art fusion filters. △ Less

Submitted 7 June, 2025; originally announced June 2025.

Complete Inertial Pose Dataset: from raw measurements to pose with low-cost and high-end MARG sensors

Authors: Manuel Palermo, Sara Cerqueira, João André, António Pereira, Cristina P. Santos

Abstract: The use of wearable technology for posture monitoring has been expanding due to its low-intrusiveness and compliance with daily use requirements. However, there are still open challenges limiting its widespread use, especially when dealing with low-cost systems. Most solutions falls either into fully functioning commercial products with high costs, or ad-hoc solutions with lower performance. Moreo… ▽ More The use of wearable technology for posture monitoring has been expanding due to its low-intrusiveness and compliance with daily use requirements. However, there are still open challenges limiting its widespread use, especially when dealing with low-cost systems. Most solutions falls either into fully functioning commercial products with high costs, or ad-hoc solutions with lower performance. Moreover, there are few datasets available, from which complete and general solutions can be derived. This work presents 2 datasets, containing low-cost and high-end Magnetic, Angular Rate, and Gravity (MARG) sensor data respectively. It provides data for the analysis of the complete inertial pose pipeline, from raw measurements, to sensor-to-segment calibration, multi-sensor fusion, skeleton kinematics, to the complete human pose. Multiple trials were collected with 21 and 10 subjects respectively, performing 6 types of sequences (ranging from calibration, to daily-activities and random movements). It presents a high degree of variability and complex dynamics with almost complete range-of-motion, while containing common sources of error found on real conditions. This amounts to 3.5M samples, synchronized with a ground-truth inertial motion capture system at 60hz. A simple end-to-end inertial pose method was briefly described and used to validate the quality of the data in both acquisitions. This database may contribute to assess, benchmark and develop novel algorithms for each of the pipelines' processing steps, with applications in classic or data-driven inertial pose estimation algorithms, human movement understanding and forecasting and ergonomic assessment in industrial or rehabilitation settings. All the data is freely available on an online database and accompanied with code to process and analyze the complete data pipeline. △ Less

Submitted 15 February, 2022; v1 submitted 12 February, 2022; originally announced February 2022.

Comments: Submitted to journal

Real-Time Human Pose Estimation on a Smart Walker using Convolutional Neural Networks

Authors: Manuel Palermo, Sara Moccia, Lucia Migliorelli, Emanuele Frontoni, Cristina P. Santos

Abstract: Rehabilitation is important to improve quality of life for mobility-impaired patients. Smart walkers are a commonly used solution that should embed automatic and objective tools for data-driven human-in-the-loop control and monitoring. However, present solutions focus on extracting few specific metrics from dedicated sensors with no unified full-body approach. We investigate a general, real-time,… ▽ More Rehabilitation is important to improve quality of life for mobility-impaired patients. Smart walkers are a commonly used solution that should embed automatic and objective tools for data-driven human-in-the-loop control and monitoring. However, present solutions focus on extracting few specific metrics from dedicated sensors with no unified full-body approach. We investigate a general, real-time, full-body pose estimation framework based on two RGB+D camera streams with non-overlapping views mounted on a smart walker equipment used in rehabilitation. Human keypoint estimation is performed using a two-stage neural network framework. The 2D-Stage implements a detection module that locates body keypoints in the 2D image frames. The 3D-Stage implements a regression module that lifts and relates the detected keypoints in both cameras to the 3D space relative to the walker. Model predictions are low-pass filtered to improve temporal consistency. A custom acquisition method was used to obtain a dataset, with 14 healthy subjects, used for training and evaluating the proposed framework offline, which was then deployed on the real walker equipment. An overall keypoint detection error of 3.73 pixels for the 2D-Stage and 44.05mm for the 3D-Stage were reported, with an inference time of 26.6ms when deployed on the constrained hardware of the walker. We present a novel approach to patient monitoring and data-driven human-in-the-loop control in the context of smart walkers. It is able to extract a complete and compact body representation in real-time and from inexpensive sensors, serving as a common base for downstream metrics extraction solutions, and Human-Robot interaction applications. Despite promising results, more data should be collected on users with impairments, to assess its performance as a rehabilitation tool in real-world scenarios. △ Less

Submitted 28 June, 2021; originally announced June 2021.

Comments: Accepted for publication in Expert Systems with Applications

Time dependence of the p/He ratio in cosmic rays according to the force-field approximation

Authors: Claudio Corti, Veronica Bindi, Cristina Consolandi, Christopher Freeman, Andrew Kuhlman, Christopher Light, Matteo Palermo, Siqi Wang

Abstract: We study the predictions for the p/He ratio in galactic cosmic rays according to the force-field approximation. The dependence of the time variation of p/He on the local interstellar spectrum (LIS) shape and on the mass-to-charge ratio, A/Z, is analyzed in detail. We find that, depending on the rigidity range and the sign of the spectral index of the p/He LIS ratio, the p/He time variation can be… ▽ More We study the predictions for the p/He ratio in galactic cosmic rays according to the force-field approximation. The dependence of the time variation of p/He on the local interstellar spectrum (LIS) shape and on the mass-to-charge ratio, A/Z, is analyzed in detail. We find that, depending on the rigidity range and the sign of the spectral index of the p/He LIS ratio, the p/He time variation can be correlated or anti-correlated with the phase of the solar cycle. We show that the A/Z dependence is the most probable cause for the p/He decrease recently observed by AMS-02 after 2015 between 2 and 3 GV. △ Less

Submitted 30 September, 2019; originally announced October 2019.

Comments: Proceeding of the 36th International Cosmic Ray Conference (ICRC19)

Test of validity of the force-field approximation with AMS-02 and PAMELA monthly fluxes

Authors: Claudio Corti, Veronica Bindi, Cristina Consolandi, Christopher Freeman, Andrew Kuhlman, Christopher Light, Matteo Palermo, Siqi Wang

Abstract: Galactic cosmic rays (GCRs) entering the heliosphere are disturbed by the magnetic field of the Sun, which varies the shape and intensity of their local interstellar spectrum. The force-field approximation is a popular way of dealing with solar modulation, especially for studies focused on galactic transport of cosmic rays. The validity of this approach to reproduce the modulated GCR fluxes at Ear… ▽ More Galactic cosmic rays (GCRs) entering the heliosphere are disturbed by the magnetic field of the Sun, which varies the shape and intensity of their local interstellar spectrum. The force-field approximation is a popular way of dealing with solar modulation, especially for studies focused on galactic transport of cosmic rays. The validity of this approach to reproduce the modulated GCR fluxes at Earth is tested using monthly proton fluxes measured by PAMELA between July 2006 and January 2014 and monthly proton and helium fluxes measured by AMS-02 between May 2011 and May 2017. We show that the precision of the new AMS-02 data requires a rigidity-dependent modification of the force-field approximation. △ Less

Submitted 30 September, 2019; originally announced October 2019.

Comments: Proceeding of the 36th International Cosmic Ray Conference (ICRC19)

Muon-induced neutrons in lead and copper at shallow depth

Authors: R. Kneißl, A. Caldwell, Q. Du, A. Empl, C. Gooch, X. Liu, B. Majorovits, M. Palermo, O. Schulz

Abstract: Next generation low-background experiments require a detailed understanding of all possible radiation backgrounds. One important radiation source are muon-induced neutrons. Their production processes are up to now not fully understood. New measurements with MINIDEX (Muon-Induced Neutron Indirect Detection EXperiment) of the production of neutrons by cosmogenic muons in high-Z materials are reporte… ▽ More Next generation low-background experiments require a detailed understanding of all possible radiation backgrounds. One important radiation source are muon-induced neutrons. Their production processes are up to now not fully understood. New measurements with MINIDEX (Muon-Induced Neutron Indirect Detection EXperiment) of the production of neutrons by cosmogenic muons in high-Z materials are reported. The setup is located at the T{ü}bingen Shallow Underground Laboratory, which provides a vertical shielding depth of (13.2\,$\pm$\,0.8) meter water equivalent at the setup location. Muon-induced neutrons are identified by the detection of 2.2\,MeV gammas from their capture on hydrogen with high-purity germanium detectors. The experimental results were compared to Geant4 Monte Carlo predictions. The measured rate of 2.2\,MeV neutron capture gammas for lead was found to be in good agreement with the Geant4 predicted rate. For copper the measured rate was found to be a factor of 0.72\,$\pm$\,0.14 lower than the Geant4 predicted rate. An additional simulation was performed using the FLUKA Monte Carlo code. The FLUKA predicted rate of detected 2.2\,MeV neutron capture gammas for lead was also found to be in good agreement with the experimental value. A detailed comparison of muon interactions and neutron production in lead for Geant4 and FLUKA revealed large discrepancies in the description of photo-nuclear and muon-nuclear inelastic scattering reactions for muon energies at shallow underground sites. These results suggest that Geant4, when used with Geant4 recommended or standard physics lists, underpredicts the neutron production in photo-nuclear inelastic scattering reactions while at the same time it overpredicts the neutron production in muon-nuclear inelastic scattering reactions. △ Less

Submitted 3 April, 2019; v1 submitted 16 January, 2019; originally announced January 2019.

Numerical modeling of galactic cosmic ray proton and helium observed by AMS-02 during the solar maximum of Solar Cycle 24

Authors: Claudio Corti, Marius S. Potgieter, Veronica Bindi, Cristina Consolandi, Chris Light, Matteo Palermo, Alexis Popkow

Abstract: Galactic cosmic rays (GCRs) are affected by solar modulation while they propagate through the heliosphere. The study of the time variation of GCR spectra observed at Earth can shed light on the underlying physical processes, specifically diffusion and particle drifts. Recently, the AMS-02 experiment measured with very high accuracy the time variation of the cosmic ray proton and helium flux betwee… ▽ More Galactic cosmic rays (GCRs) are affected by solar modulation while they propagate through the heliosphere. The study of the time variation of GCR spectra observed at Earth can shed light on the underlying physical processes, specifically diffusion and particle drifts. Recently, the AMS-02 experiment measured with very high accuracy the time variation of the cosmic ray proton and helium flux between May 2011 and May 2017 in the rigidity range from 1 to 60 GV. In this work, a comprehensive three-dimensional (3D) steady-state numerical model is used to solve Parker's transport equation and is used to reproduce the monthly proton fluxes observed by AMS-02. We find that the rigidity slope of the perpendicular mean free path above 4 GV remains constant, while below 4 GV it increases during solar maximum. Assuming the same mean free paths for helium and protons, the models are able to reproduce the time behavior of the p/He ratio observed by AMS-02. The dependence of the diffusion tensor on the particle mass-to-charge ratio, A/Z, is found to be the main cause of the time dependence of p/He below 3 GV. △ Less

Submitted 21 December, 2018; v1 submitted 22 October, 2018; originally announced October 2018.

Comments: 21 pages, 8 figures, 3 tables, accepted for publication in The Astrophysical Journal

Direct measurement of neutrons induced in lead by cosmic muons at a shallow underground site

Authors: Qiang Du, Iris Abt, Anton Empl, Chris Gooch, Raphael Kneissl, Shin Ted LIN, Bela Majorovits, Matteo Palermo, Oliver Schulz, Li Wang, Qian YUE, Anna Julia Zsigmond

Abstract: Neutron production in lead by cosmic muons has been studied with a Gadolinium doped liquid scintillator detector. The detector was installed next to the Muon-Induced Neutron Indirect Detection EXperiment (MINIDEX), permanently located in the Tübingen shallow underground laboratory where the mean muon energy is approximately 7 GeV. The MINIDEX plastic scintillators were used to tag muons; the neutr… ▽ More Neutron production in lead by cosmic muons has been studied with a Gadolinium doped liquid scintillator detector. The detector was installed next to the Muon-Induced Neutron Indirect Detection EXperiment (MINIDEX), permanently located in the Tübingen shallow underground laboratory where the mean muon energy is approximately 7 GeV. The MINIDEX plastic scintillators were used to tag muons; the neutrons were detected through neutron capture and neutron-induced nuclear recoil signals in the liquid scintillator detector. Results on the rates of observed neutron captures and nuclear recoils are presented and compared to predictions from GEANT4-9.6 and GEANT4-10.3. The predicted rates are significantly too low for both versions of GEANT4. For neutron capture events, the observation exceeds the predictions by factors of $ 1.65\,\pm\,0.02\,\textrm{(stat.)}\,\pm\,0.07\,\textrm{(syst.)} $ and $ 2.58\,\pm\,0.03\,\textrm{(stat.)}\,\pm\,0.11\,\textrm{(syst.)} $ for GEANT4-9.6 and GEANT4-10.3, respectively. For neutron nuclear recoil events, which require neutron energies above approximately 5 MeV, the factors are even larger, $ 2.22\,\pm\,0.05\,\textrm{(stat.)}\,\pm\,0.25\,\textrm{(syst.)} $ and $ 3.76\,\pm\,0.09\,\textrm{(stat.)}\,\pm\,0.41\,\textrm{(syst.)} $, respectively. Also presented is the first statistically significant measurement of the spectrum of neutrons induced by cosmic muons in lead between 5 and 40 MeV. It was obtained by unfolding the nuclear recoil spectrum. The observed neutron spectrum is harder than predicted by GEANT4. An investigation of the distribution of the time difference between muon tags and nuclear recoil signals confirms the validity of the unfolding procedure and shows that GEANT4 cannot properly describe the time distribution of nuclear recoil events. In general, the description of the data is worse for GEANT4-10.3 than for GEANT4-9.6. △ Less

Submitted 7 May, 2018; v1 submitted 15 January, 2018; originally announced January 2018.

Comments: 29 pages, 22 figures, 4 tables

Journal ref: Astroparticle Physics, Volume 102, November 2018, Pages 12-24

Alpha-event and surface characterisation in segmented true-coaxial HPGe detectors

Authors: Iris Abt, Lucia Garbini, Chris Gooch, Sabine Irlbeck, Xiang Liu, Matteo Palermo, Oliver Schulz

Abstract: A detailed study of alpha interactions on the passivation layer on the end-plate of a true-coaxial high-purity germanium detector is presented. The observation of alpha events on such a surface indicates an unexpectedly thin so-called "dead layer" of less than 20 μm thickness. In addition, the influence of the metalisation close to the end-plate on the time evolution of the output pulses is discus… ▽ More A detailed study of alpha interactions on the passivation layer on the end-plate of a true-coaxial high-purity germanium detector is presented. The observation of alpha events on such a surface indicates an unexpectedly thin so-called "dead layer" of less than 20 μm thickness. In addition, the influence of the metalisation close to the end-plate on the time evolution of the output pulses is discussed. The results indicate that alpha contamination can result in events which could be mistaken as signals for neutrinoless double beta decay and provide some guidance on how to prevent this. △ Less

Submitted 28 January, 2018; v1 submitted 1 November, 2016; originally announced November 2016.

Comments: 14 pages, 14 figures, published in Nuclear Instruments and Methods in Physics Research A

The Muon-Induced Neutron Indirect Detection EXperiment, MINIDEX

Authors: I. Abt, A. Caldwell, C. Carissimo, C. Gooch, R. Kneissl, J. Langford, X. Liu, B. Majorovits, M. Palermo, O. Schulz, L. Vanhoefer

Abstract: A new experiment to quantitatively measure neutrons induced by cosmic-ray muons in selected high-Z materials is introduced. The design of the Muon-Induced Neutron Indirect Detection EXperiment, MINIDEX, and the results from its first data taking period are presented as well as future plans. Neutron production in high-Z materials is of particular interest as such materials are used for shielding in… ▽ More A new experiment to quantitatively measure neutrons induced by cosmic-ray muons in selected high-Z materials is introduced. The design of the Muon-Induced Neutron Indirect Detection EXperiment, MINIDEX, and the results from its first data taking period are presented as well as future plans. Neutron production in high-Z materials is of particular interest as such materials are used for shielding in low-background experiments. The design of next-generation large-scale experiments searching for neutrinoless double beta decay or direct interactions of dark matter requires reliable Monte Carlo simulations of background induced by muon interactions. The first five months of operation already provided a valuable data set on neutron production and neutron transport in lead. A first round of comparisons between MINIDEX data and Monte Carlo predictions obtained with two GEANT4- based packages is presented. The rate of muon-induced events is overall a factor three to four higher in data than predicted by the Monte Carlo packages. In addition, the time evolution of the muon-induced signal is not well described by the simulations. △ Less

Submitted 5 October, 2016; originally announced October 2016.

Journal ref: Astroparticle Physics 90 (2017) 1-13

The GALATEA Test-Facility for High Purity Germanium Detectors

Authors: I. Abt, A. Caldwell, B. Doenmez, L. Garbini, S. Irlbeck, B. Majorovits, M. Palermo, O. Schulz, H. Seitz, F. Stelzer

Abstract: GALATEA is a test facility designed to investigate bulk and surface effects in high purity germanium detectors. A vacuum tank houses an infrared screened volume with a cooled detector inside. A system of three stages allows an almost complete scan of the detector. The main feature of GALATEA is that there is no material between source and detector. This allows the usage of alpha and beta sources a… ▽ More GALATEA is a test facility designed to investigate bulk and surface effects in high purity germanium detectors. A vacuum tank houses an infrared screened volume with a cooled detector inside. A system of three stages allows an almost complete scan of the detector. The main feature of GALATEA is that there is no material between source and detector. This allows the usage of alpha and beta sources as well as of a laser beam to study surface effects. A 19-fold segmented true-coaxial germanium detector was used for commissioning. △ Less

Submitted 9 February, 2015; v1 submitted 1 September, 2014; originally announced September 2014.