Showing 1–9 of 9 results for author: Soti, G

dGrasp: NeRF-Informed Implicit Grasp Policies with Supervised Optimization Slopes

Authors: Gergely Sóti, Xi Huang, Christian Wurll, Gergely Sóti

Abstract: We present dGrasp, an implicit grasp policy with an enhanced optimization landscape. This landscape is defined by a NeRF-informed grasp value function. The neural network representing this function is trained on grasp demonstrations. During training, we use an auxiliary loss to guide not only the weight updates of this network but also the update how the slope of the optimization landscape changes… ▽ More We present dGrasp, an implicit grasp policy with an enhanced optimization landscape. This landscape is defined by a NeRF-informed grasp value function. The neural network representing this function is trained on grasp demonstrations. During training, we use an auxiliary loss to guide not only the weight updates of this network but also the update how the slope of the optimization landscape changes. This loss is computed on the demonstrated grasp trajectory and the gradients of the landscape. With second order optimization, we incorporate valuable information from the trajectory as well as facilitate the optimization process of the implicit policy. Experiments demonstrate that employing this auxiliary loss improves policies' performance in simulation as well as their zero-shot transfer to the real-world. △ Less

Submitted 14 June, 2024; originally announced June 2024.

6-DoF Grasp Pose Evaluation and Optimization via Transfer Learning from NeRFs

Authors: Gergely Sóti, Xi Huang, Christian Wurll, Björn Hein

Abstract: We address the problem of robotic gras** of known and unknown objects using implicit behavior cloning. We train a grasp evaluation model from a small number of demonstrations that outputs higher values for grasp candidates that are more likely to succeed in gras**. This evaluation model serves as an objective function, that we maximize to identify successful grasps. Key to our approach is the… ▽ More We address the problem of robotic gras** of known and unknown objects using implicit behavior cloning. We train a grasp evaluation model from a small number of demonstrations that outputs higher values for grasp candidates that are more likely to succeed in gras**. This evaluation model serves as an objective function, that we maximize to identify successful grasps. Key to our approach is the utilization of learned implicit representations of visual and geometric features derived from a pre-trained NeRF. Though trained exclusively in a simulated environment with simplified objects and 4-DoF top-down grasps, our evaluation model and optimization procedure demonstrate generalization to 6-DoF grasps and novel objects both in simulation and in real-world settings, without the need for additional data. Supplementary material is available at: https://gergely-soti.github.io/grasp △ Less

Submitted 15 January, 2024; originally announced January 2024.

Gradient based Grasp Pose Optimization on a NeRF that Approximates Grasp Success

Authors: Gergely Sóti, Björn Hein, Christian Wurll

Abstract: Current robotic gras** methods often rely on estimating the pose of the target object, explicitly predicting grasp poses, or implicitly estimating grasp success probabilities. In this work, we propose a novel approach that directly maps gripper poses to their corresponding grasp success values, without considering objectness. Specifically, we leverage a Neural Radiance Field (NeRF) architecture… ▽ More Current robotic gras** methods often rely on estimating the pose of the target object, explicitly predicting grasp poses, or implicitly estimating grasp success probabilities. In this work, we propose a novel approach that directly maps gripper poses to their corresponding grasp success values, without considering objectness. Specifically, we leverage a Neural Radiance Field (NeRF) architecture to learn a scene representation and use it to train a grasp success estimator that maps each pose in the robot's task space to a grasp success value. We employ this learned estimator to tune its inputs, i.e., grasp poses, by gradient-based optimization to obtain successful grasp poses. Contrary to other NeRF-based methods which enhance existing grasp pose estimation approaches by relying on NeRF's rendering capabilities or directly estimate grasp poses in a discretized space using NeRF's scene representation capabilities, our approach uniquely sidesteps both the need for rendering and the limitation of discretization. We demonstrate the effectiveness of our approach on four simulated 3DoF (Degree of Freedom) robotic gras** tasks and show that it can generalize to novel objects. Our best model achieves an average translation error of 3mm from valid grasp poses. This work opens the door for future research to apply our approach to higher DoF grasps and real-world scenarios. △ Less

Submitted 14 September, 2023; originally announced September 2023.

Train What You Know -- Precise Pick-and-Place with Transporter Networks

Authors: Gergely Sóti, Xi Huang, Christian Wurll, Björn Hein

Abstract: Precise pick-and-place is essential in robotic applications. To this end, we define a novel exact training method and an iterative inference method that improve pick-and-place precision with Transporter Networks. We conduct a large scale experiment on 8 simulated tasks. A systematic analysis shows, that the proposed modifications have a significant positive effect on model performance. Considering… ▽ More Precise pick-and-place is essential in robotic applications. To this end, we define a novel exact training method and an iterative inference method that improve pick-and-place precision with Transporter Networks. We conduct a large scale experiment on 8 simulated tasks. A systematic analysis shows, that the proposed modifications have a significant positive effect on model performance. Considering picking and placing independently, our methods achieve up to 60% lower rotation and translation errors than baselines. For the whole pick-and-place process we observe 50% lower rotation errors for most tasks with slight improvements in terms of translation errors. Furthermore, we propose architectural changes that retain model performance and reduce computational costs and time. We validate our methods with an interactive teaching procedure on real hardware. Supplementary material will be made available at: https://gergely-soti.github.io/p △ Less

Submitted 17 February, 2023; originally announced February 2023.

Measurement of the $β$-asymmetry parameter of $^{67}$Cu in search for tensor type currents in the weak interaction

Authors: G. Soti, F. Wauters, M. Breitenfeldt, P. Finlay, P. Herzog, A. Knecht, U. Köster, I. S. Kraev, T. Porobic, P. N. Prashanth, I. S. Towner, C. Tramm, D. Zákoucký, N. Severijns

Abstract: Precision measurements at low energy search for physics beyond the Standard Model in a way complementary to searches for new particles at colliders. In the weak sector the most general $β$ decay Hamiltonian contains, besides vector and axial-vector terms, also scalar, tensor and pseudoscalar terms. Current limits on the scalar and tensor coupling constants from neutron and nuclear $β$ decay are on… ▽ More Precision measurements at low energy search for physics beyond the Standard Model in a way complementary to searches for new particles at colliders. In the weak sector the most general $β$ decay Hamiltonian contains, besides vector and axial-vector terms, also scalar, tensor and pseudoscalar terms. Current limits on the scalar and tensor coupling constants from neutron and nuclear $β$ decay are on the level of several percent. The goal of this paper is extracting new information on tensor coupling constants by measuring the $β$-asymmetry parameter in the pure Gamow-Teller decay of $^{67}$Cu, thereby testing the V-A structure of the weak interaction. An iron sample foil into which the radioactive nuclei were implanted was cooled down to milliKelvin temperatures in a $^3$He-$^4$He dilution refrigerator. An external magnetic field of 0.1 T, in combination with the internal hyperfine magnetic field, oriented the nuclei. The anisotropic $β$ radiation was observed with planar high purity germanium detectors operating at a temperature of about 10\,K. An on-line measurement of the $β$ asymmetry of $^{68}$Cu was performed as well for normalization purposes. Systematic effects were investigated using Geant4 simulations. The experimental value, $\tilde{A}$ = 0.587(14), is in agreement with the Standard Model value of 0.5991(2) and is interpreted in terms of physics beyond the Standard Model. The limits obtained on possible tensor type charged currents in the weak interaction hamiltonian are -0.045 $< (C_T+C'_T)/C_A <$ 0.159 (90\% C.L.). The obtained limits are comparable to limits from other correlation measurements in nuclear $β$ decay and contribute to further constraining tensor coupling constants. △ Less

Submitted 5 September, 2014; originally announced September 2014.

Performance of Geant4 in simulating semiconductor particle detector response in the energy range below 1 MeV

Authors: G. Soti, F. Wauters, M. Breitenfeldt, P. Finlay, I. S. Kraev, A. Knecht, T. Porobic, D. Zákoucký, N. Severijns

Abstract: Geant4 simulations play a crucial role in the analysis and interpretation of experiments providing low energy precision tests of the Standard Model. This paper focuses on the accuracy of the description of the electron processes in the energy range between 100 and 1000 keV. The effect of the different simulation parameters and multiple scattering models on the backscattering coefficients is invest… ▽ More Geant4 simulations play a crucial role in the analysis and interpretation of experiments providing low energy precision tests of the Standard Model. This paper focuses on the accuracy of the description of the electron processes in the energy range between 100 and 1000 keV. The effect of the different simulation parameters and multiple scattering models on the backscattering coefficients is investigated. Simulations of the response of HPGe and passivated implanted planar Si detectors to β particles are compared to experimental results. An overall good agreement is found between Geant4 simulations and experimental data. △ Less

Submitted 19 June, 2013; originally announced June 2013.

Hyperfine Field and Hyperfine Anomalies of Copper Impurities in Iron

Authors: V. V. Golovko, F. Wauters, S. Cottenier, M. Breitenfeldt, V. De Leebeeck, S. Roccia, G. Soti, M. Tandecki, E. Traykov, S. Van Gorp, D. Zákoucký, N. Severijns

Abstract: A new value for the hyperfine magnetic field of copper impurities in iron is obtained by combining resonance frequencies from experiments involving β-NMR on oriented nuclei on 59-Cu, 69-Cu, and 71-Cu with magnetic moment values from collinear laser spectroscopy measurements on these isotopes. The resulting value, i.e., Bhf(CuFe) = -21.794(10) T, is in agreement with the value adopted until now but… ▽ More A new value for the hyperfine magnetic field of copper impurities in iron is obtained by combining resonance frequencies from experiments involving β-NMR on oriented nuclei on 59-Cu, 69-Cu, and 71-Cu with magnetic moment values from collinear laser spectroscopy measurements on these isotopes. The resulting value, i.e., Bhf(CuFe) = -21.794(10) T, is in agreement with the value adopted until now but is an order of magnitude more precise. It is consistent with predictions from ab initio calculations. Comparing the hyperfine field values obtained for the individual isotopes, the hyperfine anomalies in Fe were determined to be 59Δ69=0.15(9)% and 71Δ69=0.07(11)%. △ Less

Submitted 29 July, 2011; originally announced July 2011.

Comments: 6 pages, 2 figures, 3 tables

Journal ref: Phys.Rev.C84:014323,2011

The half-life of $^{221}$Fr in Si and Au at 4K and at mK temperatures

Authors: F. Wauters, B. Verstichel, M. Breitenfeldt, V. De Leebeeck, V. Yu. Kozlov, I. Kraev, S. Roccia, G. Soti, M. Tandecki, E. Traykov, S. Van Gorp, D. Z'akouck'y, N. Severijns

Abstract: The half-life of the $α$ decaying nucleus $^{221}$Fr was determined in different environments, i.e. embedded in Si at 4 K, and embedded in Au at 4 K and about 20 mK. No differences in half-life for these different conditions were observed within 0.1%. Furthermore, we quote a new value for the absolute half-life of $^{221}$Fr of t$_{1/2}$ = 286.1(10) s, which is of comparable precision to the mos… ▽ More The half-life of the $α$ decaying nucleus $^{221}$Fr was determined in different environments, i.e. embedded in Si at 4 K, and embedded in Au at 4 K and about 20 mK. No differences in half-life for these different conditions were observed within 0.1%. Furthermore, we quote a new value for the absolute half-life of $^{221}$Fr of t$_{1/2}$ = 286.1(10) s, which is of comparable precision to the most precise value available in literature. △ Less

Submitted 29 October, 2010; originally announced October 2010.

Journal ref: Phys.Rev.C82:064317,2010

Precision measurements of the $^{60}$Co $β$-asymmetry parameter in search for tensor currents in weak interactions

Authors: F. Wauters, I. S. Kraev, D. Zákoucký, M. Beck, M. Breitenfeldt, V. De Leebeeck, V. V. Golovko, V. Yu. Kozlov, T. Phalet, S. Roccia, G. Soti, M. Tandecki, I. S. Towner, E. Traykov, S. Van Gorp, N. Severijns

Abstract: The $β$-asymmetry parameter $\widetilde{A}$ for the Gamow-Teller decay of $^{60}$Co was measured by polarizing the radioactive nuclei with the brute force low-temperature nuclear-orientation method. The $^{60}$Co activity was cooled down to milliKelvin temperatures in a $^3$He-$^4$He dilution refrigerator in an external 13 T magnetic field. The $β$ particles were observed by a 500 $μm$ thick Si PI… ▽ More The $β$-asymmetry parameter $\widetilde{A}$ for the Gamow-Teller decay of $^{60}$Co was measured by polarizing the radioactive nuclei with the brute force low-temperature nuclear-orientation method. The $^{60}$Co activity was cooled down to milliKelvin temperatures in a $^3$He-$^4$He dilution refrigerator in an external 13 T magnetic field. The $β$ particles were observed by a 500 $μm$ thick Si PIN diode operating at a temperature of about 10 K in a magnetic field of 0.6 T. Extensive GEANT4 Monte-Carlo simulations were performed to gain control over the systematic effects. Our result, $\widetilde{A} = -1.014(12)_{stat}(16)_{syst}$, is in agreement with the Standard-Model value of $-0.987(9)$, which includes recoil-order corrections that were addressed for the first time for this isotope. Further, it enables limits to be placed on possible tensor-type charged weak currents as well as other physics beyond the Standard Model. △ Less

Submitted 27 May, 2010; originally announced May 2010.

Journal ref: Phys.Rev.C82:055502,2010