Searcharxiv⌕ Search

arXiv subjects

Michael Lorenz

Publications and source records attributed to Michael Lorenz.

10 recordsLinked to original sources

Amortized Inverse Kinematics via Graph Attention for Real-Time Human Avatar Animation

Inverse kinematics (IK) is a core operation in animation, robotics, and biomechanics: given Cartesian constraints, recover joint rotations under a known kinematic tree. In many real-time human avatar pipelines, the available signal per frame is a sparse set of tracked 3D joint positions, whereas animation systems require joint orientations to drive skinning. Recovering full orientations from positions is underconstrained, most notably because twist about bone axes is ambiguous, and classical IK solvers typically rely on iterative optimization that can be slow and sensitive to noisy inputs. We introduce IK-GAT, a lightweight graph-attention network that reconstructs full-body joint orientations from 3D joint positions in a single forward pass. The model performs message passing over the skeletal parent-child graph to exploit kinematic structure during rotation inference. To simplify learning, IK-GAT predicts rotations in a bone-aligned world-frame representation anchored to rest-pose bone frames. This parameterization makes the twist axis explicit and is exactly invertible to standard parent-relative local rotations given the kinematic tree and rest pose. The network uses a continuous 6D rotation representation and is trained with a geodesic loss on SO(3) together with an optional forward-kinematics consistency regularizer. IK-GAT produces animation-ready local rotations that can directly drive a rigged avatar or be converted to pose parameters of SMPL-like body models for real-time and online applications. With 374K parameters and over 650 FPS on CPU, IK-GAT outperforms VPoser-based per-frame iterative optimization without warm-start at significantly lower cost, and is robust to initial pose and input noise

cs.CV↗

Evaluating Large Language Models for Gait Classification Using Text-Encoded Kinematic Waveforms

Background: Machine learning (ML) enhances gait analysis but often lacks the level of interpretability desired for clinical adoption. Large Language Models (LLMs) may offer explanatory capabilities and confidence-aware outputs when applied to structured kinematic data. This study therefore evaluated whether general-purpose LLMs can classify continuous gait kinematics when represented as textual numeric sequences and how their performance compares to conventional ML approaches. Methods: Lower-body kinematics were recorded from 20 participants performing seven gait patterns. A supervised KNN classifier and a class-independent One-Class SVM (OCSVM) were compared against zero-shot LLMs (GPT-5, GPT-5-mini, GPT-4.1, and o4-mini). Models were evaluated using Leave-One-Subject-Out (LOSO) cross-validation. LLMs were tested both with and without explicit reference gait statistics. Results: The supervised KNN achieved the highest performance (multiclass Matthews Correlation Coefficient, MCC = 0.88). The best-performing LLM (GPT-5) with reference grounding achieved a multiclass MCC of 0.70 and a binary MCC of 0.68, outperforming the class-independent OCSVM (binary MCC = 0.60). Performance of the LLM was highly dependent on explicit reference information and self-rated confidence; when restricted to high-confidence predictions, multiclass MCC increased to 0.83 on the filtered subset. Notably, the computationally efficient o4-mini model performed comparably to larger models. Conclusion: When continuous kinematic waveforms were encoded as textual numeric tokens, general-purpose LLMs, even with reference grounding, did not match supervised multiclass classifiers for precise gait classification and are better regarded as exploratory systems requiring cautious, human-guided interpretation rather than diagnostic use.

cs.LG↗

MinJointTracker: Real-time inertial kinematic chain tracking with joint position estimation and minimal state size

Inertial motion capture is a promising approach for capturing motion outside the laboratory. However, as one major drawback, most of the current methods require different quantities to be calibrated or computed offline as part of the setup process, such as segment lengths, relative orientations between inertial measurement units (IMUs) and segment coordinate frames (IMU-to-segment calibrations) or the joint positions in the IMU frames. This renders the setup process inconvenient. This work contributes to real-time capable calibration-free inertial tracking of a kinematic chain, i.e. simultaneous recursive Bayesian estimation of global IMU angular kinematics and joint positions in the IMU frames, with a minimal state size. Experimental results on simulated IMU data from a three-link kinematic chain (manipulator study) as well as re-simulated IMU data from healthy humans walking (lower body study) show that the calibration-free and lightweight algorithm provides not only drift-free relative but also drift-free absolute orientation estimates with a global heading reference for only one IMU as well as robust and fast convergence of joint position estimates in the different movement scenarios.

cs.RO↗

Impact of magnetization and hyperfine field distribution on high magnetoelectric coupling strength in BaTiO$_3$-BiFeO$_3$ multilayers

Understanding the mechanisms of magnetoelectric (ME) coupling within multiferroic structures is paramount from a fundamental as well as an applied point of view. We report here that the magnetoelectric properties, as well as the magnetization, of BaTiO$_3$-BiFeO$_3$ superlattices can be tuned by varying the BiFeO$_3$ layer thickness. The magnetoelectric voltage coefficient ($α_{ME}$) reaches its maximum of 60.2 Vcm$^{-1}$Oe$^{-1}$ at 300 K, one of the highest values reported so far, for a sample with a BiFeO$_3$ thickness of 5 nm and a BaTiO$_3$ thickness of 10 nm. To gain deeper insight into the increased magnetoelectric coupling, and both the local and macroscopic magnetic properties, samples with varying BiFeO$_3$ thicknesses have been investigated. Correlations were established between the hyperfine field (HFF), the magnetoelectric voltage coefficient and the magnetization. The possible mechanisms responsible for the strong magnetoelectric coupling are discussed.

cond-mat.mtrl-sci↗

Effect of annealing on the magnetic properties of zinc ferrite thin films

We report on the magnetic properties of zinc ferrite thin film deposited on SrTiO$_3$ single crystal using pulsed laser deposition. X-ray diffraction result indicates the highly oriented single phase growth of the film along with the presence of the strain. In comparison to the bulk antiferromagnetic order, the as-deposited film has been found to exhibit ferrimagnetic ordering with a coercive field of 1140~Oe at 5~K. A broad maximum, at $\approx$105~K, observed in zero-field cooled magnetization curve indicates the wide grain size distribution for the as-deposited film. Reduction in magnetization and blocking temperature has been observed after annealing in both argon as well as oxygen atmospheres, where the variation was found to be dependent on the annealing temperature.

cond-mat.mtrl-sci↗

Fundamental absorption edges in heteroepitaxial YBiO$_3$ thin films

The dielectric function of heteroepitaxial YBiO$_3$ grown on $a$-Al$_2$O$_3$ single crystals via pulsed laser deposition is determined in the spectral range from 0.03 eV to 4.5 eV by simultaneous modeling of spectroscopic ellipsometry and optical transmission data of YBiO$_3$ films of different thickness. The (111)-oriented YBiO$_3$ films are nominally unstrained and crystallize in a defective fluorite-type structure with $Fm\bar{3}m$ space group. From the calculated absorption spectrum, a direct electronic bandgap energy of 3.6(1) eV and the signature of an indirect electronic transition around 0.5 eV are obtained. These values provide necessary experimental feedback to previous conflicting electronic band structure calculations predicting either a topologically trivial or non-trivial insulating ground state in YBiO$_3$.

cond-mat.mtrl-sci↗

Comparative Study of Optical and Magneto-Optical Properties of Normal, Disordered and Inverse Spinel Type Oxides

Co$_3$O$_4$, ZnFe$_2$O$_4$, CoFe$_2$O$_4$, ZnCo$_2$O$_4$, and Fe$_3$O$_4$ thin films were fabricated by pulsed laser deposition at high and low temperatures resulting in crystalline single-phase normal, inverse, as well as disordered spinel oxide thin films with smooth surface morphology. The dielectric function, determined by spectroscopic ellipsometry in a wide spectral range from 0.5 eV to 8.5 eV, is compared with the magneto-optical response of the dielectric tensor, investigated by magneto-optical Kerr effect (MOKE) spectroscopy in the spectral range from 1.7 eV to 5.5 eV with an applied magnetic field of 1.7 T. Crystal field, inter-valence and inter-sublattice charge transfer transitions, and transitions from O$_{2p}$ to metal cation 3d or 4s bands are identified in both the principal diagonal elements and the magneto-optically active off-diagonal elements of the dielectric tensor. Depending on the degree of cation disorder, resulting in local symmetry distortion, the magneto-optical response is found to be strongest for high crystal quality inverse spinels and for disordered normal spinel structure, contrary to the first principle studies of CoFe$_2$O$_4$ and ZnFe$_2$O$_4$. The results presented provide a basis for deeper understanding of light-matter interaction in this material system that is of vital importance for device-related phenomena and engineering.

cond-mat.mtrl-sci↗

Electronic excitations and structure of Li2IrO3 thin films grown on ZrO$_{2}$:Y (001) substrates

Thin films are a prerequisite for application of the emergent exotic ground states in iridates that result from the interplay of strong spin-orbit coupling and electronic correlations. We report on pulsed laser deposition of Li$_{2}$IrO$_{3}$ films on ZrO$_{2}$:Y(001) single crystalline substrates. X-ray diffraction confirms preferential (001) and (10-1) out-of-plane crystalline orientations with well defined in-plane orientation. Resistivity between 35 and 300 K is dominated by a three-dimensional variable range hopping mechanism. The dielectric function is determined by means of spectroscopic ellipsometry and, complemented by Fourier transform infrared transmission spectroscopy, reveals a small optical gap of $\approx$ 300 meV, a splitting of the $5d$-$t_{2g}$ manifold, and several in-gap excitations attributed to phonons and possibly magnons.

cond-mat.str-el↗

Mott Variable Range Hopping and Weak Antilocalization Effect in Heteroepitaxial Na2IrO3 Thin Films

Iridate thin films are a prerequisite for any application utilizing their cooperative effects resulting from the interplay of stron spin-orbit coupling and electronic correlations. Here, heteroepitaxial Na2IrO3 thin films with excellent (001) out-of-plane crystalline orientation and well defined in-plane epitaxial relationship are presented on various oxide substrates. Resistivity is dominated by a three-dimensional variable range hopping mechanism in a large temperature range between 300 K and 40 K. Optical experiments show the onset of a small optical gap of about 200 meV and a splitting of the Ir 5d-t2g manifold. Positive magnetoresistance below 3 T and 25 K shows signatures of a weak antilocalization effect. This effect can be associated with surface states in a topological insulator and hence supports proposals for a topological insulator phase present in Na2IrO3.

cond-mat.str-el↗

Room temperature ferromagnetism in carbon-implanted ZnO

Unexpected ferromagnetism has been observed in carbon doped ZnO films grown by pulsed laser deposition [Phys. Rev. Lett. 99, 127201 (2007)]. In this letter, we introduce carbon into ZnO films by ion implantation. Room temperature ferromagnetism has been observed. Our analysis demonstrates that (1) C-doped ferromagnetic ZnO can be achieved by an alternative method, i.e. ion implantation, and (2) the chemical involvement of carbon in the ferromagnetism is indirectly proven.

cond-mat.mtrl-sci↗