arXiv subjects

Junfan Zhu

Publications and source records attributed to Junfan Zhu.

7 recordsLinked to original sources

arXiv2026

GestureLSM: Latent Shortcut based Co-Speech Gesture Generation with Spatial-Temporal Modeling

Generating full-body human gestures based on speech signals remains challenges on quality and speed. Existing approaches model different body regions such as body, legs and hands separately, which fail to capture the spatial interactions between them and result in unnatural and disjointed movements. Additionally, their autoregressive/diffusion-based pipelines show slow generation speed due to dozens of inference steps. To address these two challenges, we propose GestureLSM, a flow-matching-based approach for Co-Speech Gesture Generation with spatial-temporal modeling. Our method i) explicitly model the interaction of tokenized body regions through spatial and temporal attention, for generating coherent full-body gestures. ii) introduce the flow matching to enable more efficient sampling by explicitly modeling the latent velocity space. To overcome the suboptimal performance of flow matching baseline, we propose latent shortcut learning and beta distribution time stamp sampling during training to enhance gesture synthesis quality and accelerate inference. Combining the spatial-temporal modeling and improved flow matching-based framework, GestureLSM achieves state-of-the-art performance on BEAT2 while significantly reducing inference time compared to existing methods, highlighting its potential for enhancing digital humans and embodied agents in real-world applications. Project Page: https://andypinxinliu.github.io/GestureLSM

cs.CV↗

arXiv2026

SSP: An Event-Matched Syn2Sim2Phy Cross-Domain Evaluation Framework for Autonomous Driving VLA Models

Vision-language-action (VLA) models for autonomous driving jointly produce scene interpretation, language-based reasoning, and driving trajectories. Existing evaluations often use independently selected synthetic, simulated, and physical data, so measured performance gaps can be confounded by changes in scenario content rather than genuine domain sensitivity. We propose SSP (Synthetic-Simulation-Physical), an event-matched Syn2Sim2Phy evaluation framework that anchors cross-domain comparison to the same safety-critical interaction. Starting from a synthetic long-tail video, SSP builds a validated event specification that preserves road topology, participant roles, relative motion, conflict evolution, passing order, response constraints, and event phases. Platform-specific realizations are then constructed in CARLA and on a closed proving ground and are evaluated only after transfer audits confirm preservation of mandatory event properties. SSP maps heterogeneous outputs from OpenEMMA, LLaViDA, and Alpamayo-R1 into common semantic slots and a 1 s trajectory window to assess output validity, semantic accuracy, critical-interaction recognition, trajectory quality, and risk response. Across Cut-in and vulnerable-road-user crossing cases, the macro-averaged Integrated VLA Capability Scores are 0.259, 0.291, and 0.325 in the Synthetic, Simulation, and Physical domains, respectively, while the best domain varies by scenario. Alpamayo-R1, OpenEMMA, and LLaViDA obtain scores of 0.405, 0.338, and 0.131. SSP provides a reproducible scene-transfer chain and an evidence-qualified evaluation of VLA behavior without assuming that the Physical domain is universally superior.

cs.CV↗

arXiv2026

An interactive enhanced driving dataset for autonomous driving

Driving interaction data are important for training and evaluating autonomous drivingVision-Language-Action (VLA) models, but existing datasets contain limited denseinteraction samples and weak alignment between trajectories, visual inputs, and languageannotations. This work presents the Interactive Enhanced Driving Dataset (IEDD), alarge-scale interaction-oriented dataset constructed from five naturalistic trajectory datasets:Lyft Level 5, Waymo, nuPlan, INTERACTION, and SIND. IEDD contains 7.31 millionego-centric interaction segments, including 6.66 million multi agents cases, covering head-on,car-following, merging, and crossing interactions. Each segment is associated withtrajectory-derived interaction metrics describing interaction intensity and efficiency. Based onthese annotations, IEDD-VQA further provides trajectory-reconstructed BEV videos,structured interaction semantics, and multi-turn question-answer pairs. The dataset cansupport interaction mining, long-tail scenario analysis, VLA instruction tuning, andhierarchical evaluation of perception, behavior description, physical quantification, andcounterfactual reasoning.

cs.CV↗

arXiv2026

QuantEval: A Benchmark for Financial Quantitative Tasks in Large Language Models

Large Language Models (LLMs) have shown strong capabilities across many domains, yet their evaluation in financial quantitative tasks remains fragmented and mostly limited to knowledge-centric question answering. We introduce QuantEval, a benchmark that evaluates LLMs across three essential dimensions of quantitative finance: knowledge-based QA, quantitative mathematical reasoning, and quantitative strategy coding. Unlike prior financial benchmarks, QuantEval integrates a CTA-style backtesting framework that executes model-generated strategies and evaluates them using financial performance metrics, enabling a more realistic assessment of quantitative coding ability. We evaluate some state-of-the-art open-source and proprietary LLMs and observe substantial gaps to human experts, particularly in reasoning and strategy coding. Finally, we conduct large-scale supervised fine-tuning and reinforcement learning experiments on domain-aligned data, demonstrating consistent improvements. We hope QuantEval will facilitate research on LLMs' quantitative finance capabilities and accelerate their practical adoption in real-world trading workflows. We additionally release the full deterministic backtesting configuration (asset universe, cost model, and metric definitions) to ensure strict reproducibility.

cs.CL↗

arXiv2025

Single-Light-Pulse Driven Compact Atom Interferometry with Measurement Induced Large Momentum Transfer

We propose a fundamentally new design strategy of light-pulsed atom interferometry (LPAI) with a single atomic beam splitter. A traditional $π/2$-pulse Raman beam is employed to render a small momentum transfer at the initial state. After a short period of evolution during which physical relevant information can be loaded, a quantum weak measurement is applied to the internal state of the atoms. The final information will be detected from the transmission spectrum of a probe light to obviate the measurement of florescence signal. An effective amplification of the order of $10^3$ about the momentum offset is achieved in our simulation employing $Cs$ atoms with current experimental condition. Our proposal offers a cost-effective, high-accuracy measurement and readout strategy for LPAI. Furthermore, the strategy makes the physical setup much simpler and more compact offering new direction towards portable sensitive LPAI.

quant-ph↗

arXiv2024

Topologically protected measurement of orbital angular momentum of light

We develop a weak measurement scheme for measuring orbital angular momentum (OAM) of light based on the global topology in wave function. We introduce the spin-orbit coupling to transform the measurement of OAM to the pre- and postselected measurement of polarization. The OAM number can be precisely and promptly recognized using single-shot detection without the need for spatial resolution. More significantly, the measurement results exhibit topological robustness under random phase perturbations. This scheme has the potential to be applied as a paradigm in the OAM-based optical computing, metrology and communication.

physics.optics↗

arXiv2022

A general scheme of differential imaging employing weak measurement

We propose and experimentally realize a general scheme of differential imaging employing the idea of weak measurement. We show that the weak coupling between the system of interest and a two-level ancilla can introduce a two-beam circuit after an arbitrary pre-selection of the ancilla. By choosing the post-selection orthogonal to the pre-selection measurement, an effective imaging platform based on differential operations is shown achieved. Experimental results on both the Sagnac interferometer and ultra-thin Wollaston prism demonstrate that our imaging scheme successfully yields the boundary information of complex geometric configurations.

physics.optics↗