Searcharxiv⌕ Search

arXiv subjects

Hongyu Ma

Publications and source records attributed to Hongyu Ma.

7 recordsLinked to original sources

WinDEX: Metric 3D Hand Motion Reconstruction in the Wild for Dexterous Teleoperation and Policy Learning

Human manipulation data are widely used to train robots for human-like dexterity. However, low-cost, large-scale collection in the wild remains challenging. Existing methods remain limited in geometrically grounded metric-scale reconstruction, real-world generalization and inference speed. We present WinDEX, integrating (1) geometric-temporal optimization for metric 3D hand annotation from in-the-wild egocentric stereo videos, combined with visual-inertial SLAM for world-frame reconstruction; (2) WinDEX-Net, an end-to-end, fisheye-native network combining calibrated stereo, local detail and temporal context, with a data-model flywheel of model-assisted annotation and retraining; (3) distilled models for offline reconstruction and causal streaming, supporting reinforcement learning and teleoperation; and (4) low-cost CyberCap stereo-inertial capture, forming the WinDEX-Hand dataset with 1,000 hours of training annotations and held-out human-annotated in-the-wild 2D and independent optical motion-capture evaluation sets. Offline reconstruction runs at 43.61 FPS and causal streaming at 171 FPS. On held-out data, offline reconstruction achieves state-of-the-art accuracy: 14.5 pixels in-the-wild reprojection error and, with visual-inertial odometry, 37.29 mm world-frame joint error. We achieve 7.6 and 3.6 times the task success rates of reinforcement learning policies trained with WiLoR and HaWoR motion references, respectively, and demonstrate real-time arm-hand teleoperation. Project website: https://fruitrabbit.github.io/WinDEX/.

cs.RO↗

ESTHER: Egocentric Stereo Hand Estimation and Reconstruction in the Wild

Human dexterity is guided by two eyes watching two hands: binocular vision supplies the metric 3D structure that fine-grained manipulation consumes. Egocentric stereo is therefore the natural perceptual interface for robots, AR, and VR-yet metric 3D hand reconstruction from this very signal still has neither an end-to-end model nor an in-the-wild benchmark. We propose ESTHER, a model whose stereo geometry, temporal reasoning, and output representation are designed for wearable egocentric stereo. It is trained on pseudo-labels from a calibrated labeling pipeline and in turn assembles our benchmark ESTHER3D, an egocentric stereo hand dataset pairing a large in-the-wild training set of model-generated labels with a motion capture test set of true metric ground truth. Experiments show state-of-the-art accu?racy, superior external generalization, and robustness to the missing views, dropped frames, and lighting and motion blur extremes of real egocentric capture that break existing meth?ods. This robustness runs deeper than graceful degradation: stereo guidance teaches the model to bind apparent hand scale to metric depth, so it not only adapts to different stereo rigs and modalities with minimal fine-tuning, but more strikingly preserves true metric scale even after collapsing to a single monocular view.

cs.CV↗

Information-preserving boundary reconstruction for center-flux energetics in a regulated SU(3) gauge ladder

Gauss constraints make it nontrivial to open, close or join a gauge system while retaining its interior physics. Boundary reconstructions are constructed on a pure SU(3) ladder at six site-singlet cutoffs, with all allowed intertwiner multiplicities. Replacements of prescribed length are obtained using legal column paths and physical self-loops. On their stated physical domains, the channels preserve unnormalized conditional interior matrices, including their weights and within-branch coherence. Lowest-cutoff dephasing controls show that path probabilities alone do not preserve the matched magnetic observables. In that example, coherent resources lower the output energy without purification or global ground-state preparation. Matched interior Hamiltonian terms cancel exactly, so the energy cost is bounded independently of longitudinal length. The vacuum and flux energy densities are then established separately by same-sector gluing and stability. The open and periodic vacuum-subtracted line-energy densities are shown to be equal by two same-length boundary comparisons using the qualified periodic minimizing domain. These limits hold at fixed lattice spacing, transverse width, coupling and chosen cutoff.

hep-lat↗

Metallic Contact Contributions in Thermal Hall Conductivity Measurements

We investigate the influence of metallic contacts on thermal Hall measurements. By analyzing typical measurement setups, we show that heat currents bypassing through metallic contacts could generate non-negligible thermal Hall signals. We find that contributions from metallic contacts with thicknesses on the order of 10$^{-2}$ of sample widths can approximately replicate experimental observations across different materials in both temperature dependence and magnitude, assuming silver contacts with a conductivity of $10^{8}~\mathrm{S/m}$. Our analysis underscores the need to minimize metallic contact effects in thermal Hall measurements, which can be achieved by optimizing measurement configurations.

cond-mat.mtrl-sci↗

Exploring Graph-based Knowledge: Multi-Level Feature Distillation via Channels Relational Graph

In visual tasks, large teacher models capture essential features and deep information, enhancing performance. However, distilling this information into smaller student models often leads to performance loss due to structural differences and capacity limitations. To tackle this, we propose a distillation framework based on graph knowledge, including a multi-level feature alignment strategy and an attention-guided mechanism to provide a targeted learning trajectory for the student model. We emphasize spectral embedding (SE) as a key technique in our distillation process, which merges the student's feature space with the relational knowledge and structural complexities similar to the teacher network. This method captures the teacher's understanding in a graph-based representation, enabling the student model to more accurately mimic the complex structural dependencies present in the teacher model. Compared to methods that focus only on specific distillation areas, our strategy not only considers key features within the teacher model but also endeavors to capture the relationships and interactions among feature sets, encoding these complex pieces of information into a graph structure to understand and utilize the dynamic relationships among these pieces of information from a global perspective. Experiments show that our method outperforms previous feature distillation methods on the CIFAR-100, MS-COCO, and Pascal VOC datasets, proving its efficiency and applicability.

cs.CV↗

Evolution of Spectra for Mechanical and Wind Waves in a Large Tank

Empirical spectra for mechanical and wind waves measured in a large tank of the First Institute of Oceanography of China are presented. Analysis for the first and the second type of waves is done separately. It is shown that, in the case of mechanical waves with a steepness more than 0.2, the frequency spectra of waves evolve to ones with the tail decay S(f) ~ (f to power -4.2), whilst the shape of spectra at large fetches is self-similar. Numerical solutions of the four-wave kinetic equation, written in the fetch-limited version, result in the same spectra. This allows treating the empirical observations for mechanical waves as the natural evolution of free nonlinear surface waves. In the case of wind waves, the wave spectra evolve to ones with the tail decay S(f) ~ (f to power -4.0) at fetches X greater than 8 meters, under any applied winds. The well-known Toba three-second relation, H ~ (T to power 3/2), between the mean wave height, H, and the peak period, T, is well fulfilled. Though, the intensities of spectra tails do not follow the Toba ratio, S(f) ~ (gu)(f to power -4), rather they better correspond to the ratio , S(f) ~ ((u to power 2)/Xg)(to power p)(gu)(f to power -4) with p = 1/3. Moreover, the shape of the wind-wave spectra is not self-similar. Discussion of the results is presented.

physics.ao-ph↗

Shear-Exfoliated Phosphorene for Rechargeable Nanoscale Battery

Discovery of atomically thin black phosphorus (called phosphorene) holds promise to be used as an alternative two-dimensional material to graphene and transition metal dichalcogenides especially as an anode material for lithium-ion batteries (LIBs). However, at present bulk black phosphorus (BP) still suffers from rapid capacity fading that results in poor rechargeable performance. Here, for the first time, we use in situ transmission electron microscopy (TEM) to construct nanoscale phosphorene LIBs and visualize the capacity fading mechanism in thick multilayer phosphorene by real time capturing delithiation-induced structural decomposition that reduces electrical conductivity and thus causes irreversibility of lithiated Li3P phase. We further demonstrate that few-layer phosphorene successfully circumvents the structural decomposition and holds superior structural restorability, even subjected to multi-cycle lithiation/delithiation processes and concomitant huge volume expansion. This finding affords new experimental insights into thickness-dependent lithium diffusion kinetics in phosphorene. Additionally, a scalable liquid-phase shear exfoliation route has been developed to produce high-quality ultrathin (monolayer or few-layer) phosphorene, only by a high-speed shear mixer or even a household kitchen blender with the shear rate threshold, which will pave the way for potential large-scale applications in LIBs once the rechargeable phosphorene nanoscale batteries can be transferred to industrialized enlargement in the future.

cond-mat.mtrl-sci↗