arXiv subjects

Siyuan Fan

Publications and source records attributed to Siyuan Fan.

6 recordsLinked to original sources

arXiv2026

Structure-preserving operator splitting for 2.5D ideal MHD with an entropy-stable DGSEM and exactly divergence-free compatible finite elements

We develop a high-order, fully explicit operator-splitting method for 2.5D ideal magnetohydrodynamics on Cartesian meshes. The hydrodynamic subflow is advanced by an entropy-stable discontinuous Galerkin spectral element method equipped with stagewise oscillation elimination and a positivity-preserving limiter. The magnetic--velocity subflow uses compatible finite elements and mass-lumped reconstructions of electric field and current density. Its discrete-curl update exactly preserves the global $H(\mathrm{div})$ divergence-free subspace, while the ideal semidiscretization balances kinetic, magnetic, and internal energy. Curl-form artificial resistivity and a direction-resolved velocity filter return removed magnetic and kinetic energy to internal energy; consequently, the stabilized magnetic stage preserves positive internal energy and satisfies a discrete entropy inequality. The two solvers are composed by a second-order hydrodynamic--magnetic--hydrodynamic Strang splitting, yielding a matrix-free scheme that retains global mass, nodal admissibility, and magnetic divergence on accepted steps. Smooth Alfv\'en-wave and advected-vortex tests reveal an even--odd convergence pattern in the magnetic polynomial degree, verify second-order temporal accuracy, and show that stabilization preserves high-order accuracy. Field-loop, Orszag--Tang, rotor, MHD blast-wave, and Kelvin--Helmholtz calculations demonstrate robust performance for nonsmooth multidimensional flows and show that artificial resistivity suppresses grid-scale magnetic oscillations while retaining the resolved structures.

math.NA↗

arXiv2026

On Aligning Hierarchical Standardized Embedding for Audio-visual Generalized Zero-shot Learning

Audio-visual Generalized Zero-shot Learning (AV-GZSL) is a challenging task that aims to classify both seen and unseen objects or scenes by integrating data from audio and visual modalities. Recent studies primarily focus on fusing or aligning audio and visual features to generate more informative audio-visual embeddings. Also, aligning the audio-visual and textual features of most existing methods relies solely on the optimization objectives. However, those methods neglect the inherent distributional and structural differences between audio-visual and textual modalities. To address this limitation, we propose a method termed Aligning Hierarchical Standardized Embedding (AHSE), which enables hierarchical alignment of standardized audio-visual and textual embeddings within a shared embedding space. Specifically, we first apply Z-score standardization to the fused audio-visual and textual embeddings to reduce distributional mismatches. We then introduce a hierarchical alignment strategy that minimizes discrepancies at the semantic, class, and batch levels, thereby constructing a more robust and well-structured embedding space. This strategy not only preserves semantic and inter-class relationships but also maintains spatial consistency within each batch. Extensive experiments on three benchmark datasets: VGGSound-GZSL, UCF-GZSL, and ActivityNet-GZSL, demonstrate that AHSE achieves competitive performance in zero-shot learning.

cs.CV↗

arXiv2026

Generalized Multidimensional Contests with Asymmetric Players: Equilibrium and Optimal Prize Design

We study $n$-dimensional contests between two players with heterogeneous effort costs, where each dimension (battle) is modeled as a Tullock contest. Prize-allocation rules are identity-independent, budget-balanced, and weakly increasing in the number of victories. Players' costs can be separable across battles or exhibit cross-battle externalities. We identify a tight sufficient condition under which a unique equilibrium exists and is in pure strategies, for all admissible prize-allocation rules and all degrees of player asymmetry. Under this condition, we characterize the effort-maximizing prize-allocation rule: the entire prize goes to the player who wins more battles than the opponent by at least a prespecified margin, and is split equally if neither player meets this threshold. In the symmetric-player case, the majority rule is optimal if $n$ is odd. Interestingly, cross-battle cost externalities do not change the optimal prize allocation rule in our setting.

econ.TH↗

arXiv2025

EmbodiedBrain: Expanding Performance Boundaries of Task Planning for Embodied Intelligence

The realization of Artificial General Intelligence (AGI) necessitates Embodied AI agents capable of robust spatial perception, effective task planning, and adaptive execution in physical environments. However, current large language models (LLMs) and multimodal LLMs (MLLMs) for embodied tasks suffer from key limitations, including a significant gap between model design and agent requirements, an unavoidable trade-off between real-time latency and performance, and the use of unauthentic, offline evaluation metrics. To address these challenges, we propose EmbodiedBrain, a novel vision-language foundation model available in both 7B and 32B parameter sizes. Our framework features an agent-aligned data structure and employs a powerful training methodology that integrates large-scale Supervised Fine-Tuning (SFT) with Step-Augumented Group Relative Policy Optimization (Step-GRPO), which boosts long-horizon task success by integrating preceding steps as Guided Precursors. Furthermore, we incorporate a comprehensive reward system, including a Generative Reward Model (GRM) accelerated at the infrastructure level, to improve training efficiency. For enable thorough validation, we establish a three-part evaluation system encompassing General, Planning, and End-to-End Simulation Benchmarks, highlighted by the proposal and open-sourcing of a novel, challenging simulation environment. Experimental results demonstrate that EmbodiedBrain achieves superior performance across all metrics, establishing a new state-of-the-art for embodied foundation models. Towards paving the way for the next generation of generalist embodied agents, we open-source all of our data, model weight, and evaluating methods, which are available at https://zterobot.github.io/EmbodiedBrain.github.io.

cs.CV↗

arXiv2025

3D Human Interaction Generation: A Survey

3D human interaction generation has emerged as a key research area, focusing on producing dynamic and contextually relevant interactions between humans and various interactive entities. Recent rapid advancements in 3D model representation methods, motion capture technologies, and generative models have laid a solid foundation for the growing interest in this domain. Existing research in this field can be broadly categorized into three areas: human-scene interaction, human-object interaction, and human-human interaction. Despite the rapid advancements in this area, challenges remain due to the need for naturalness in human motion generation and the accurate interaction between humans and interactive entities. In this survey, we present a comprehensive literature review of human interaction generation, which, to the best of our knowledge, is the first of its kind. We begin by introducing the foundational technologies, including model representations, motion capture methods, and generative models. Subsequently, we introduce the approaches proposed for the three sub-tasks, along with their corresponding datasets and evaluation metrics. Finally, we discuss potential future research directions in this area and conclude the survey. Through this survey, we aim to offer a comprehensive overview of the current advancements in the field, highlight key challenges, and inspire future research works.

cs.CV↗

arXiv2024

TextIM: Part-aware Interactive Motion Synthesis from Text

In this work, we propose TextIM, a novel framework for synthesizing TEXT-driven human Interactive Motions, with a focus on the precise alignment of part-level semantics. Existing methods often overlook the critical roles of interactive body parts and fail to adequately capture and align part-level semantics, resulting in inaccuracies and even erroneous movement outcomes. To address these issues, TextIM utilizes a decoupled conditional diffusion framework to enhance the detailed alignment between interactive movements and corresponding semantic intents from textual descriptions. Our approach leverages large language models, functioning as a human brain, to identify interacting human body parts and to comprehend interaction semantics to generate complicated and subtle interactive motion. Guided by the refined movements of the interacting parts, TextIM further extends these movements into a coherent whole-body motion. We design a spatial coherence module to complement the entire body movements while maintaining consistency and harmony across body parts using a part graph convolutional network. For training and evaluation, we carefully selected and re-labeled interactive motions from HUMANML3D to develop a specialized dataset. Experimental results demonstrate that TextIM produces semantically accurate human interactive motions, significantly enhancing the realism and applicability of synthesized interactive motions in diverse scenarios, even including interactions with deformable and dynamically changing objects.

cs.CV↗