arXiv subjects

Haiyang Liu

Publications and source records attributed to Haiyang Liu.

At least 37 records · Page 2Linked to original sources

arXiv2025

A persistent-homology-based Bayesian prior for potential coefficient reconstruction in an elliptic PDE

We address the reconstruction of a potential coefficient in an elliptic partial differential equation from distributed observations within the Bayesian framework. The choice of prior distribution is crucial in such inverse problems, particularly when the target function exhibits sharp discontinuities that conventional Gaussian priors fail to capture effectively. To overcome this limitation, we introduce a novel prior based on persistent homology (PH), which quantifies and encodes the topological features of candidate functions through their persistent pairs. To ensure a well-defined distribution in infinite-dimensional spaces, the prior is constructed with respect to a Gaussian reference measure. A significant advantage over classical approaches is that the PH prior only requires the unknown functions to belong to a suitable topological space, which substantially enhances its applicability. Numerical results demonstrate that the proposed PH prior outperforms the Gaussian prior and achieves a modest yet consistent improvement over the classical total variation (TV) prior.

math.NA↗

arXiv2025

Droplet3D: Commonsense Priors from Videos Facilitate 3D Generation

Scaling laws have validated the success and promise of large-data-trained models in creative generation across text, image, and video domains. However, this paradigm faces data scarcity in the 3D domain, as there is far less of it available on the internet compared to the aforementioned modalities. Fortunately, there exist adequate videos that inherently contain commonsense priors, offering an alternative supervisory signal to mitigate the generalization bottleneck caused by limited native 3D data. On the one hand, videos capturing multiple views of an object or scene provide a spatial consistency prior for 3D generation. On the other hand, the rich semantic information contained within the videos enables the generated content to be more faithful to the text prompts and semantically plausible. This paper explores how to apply the video modality in 3D asset generation, spanning datasets to models. We introduce Droplet3D-4M, the first large-scale video dataset with multi-view level annotations, and train Droplet3D, a generative model supporting both image and dense text input. Extensive experiments validate the effectiveness of our approach, demonstrating its ability to produce spatially consistent and semantically plausible content. Moreover, in contrast to the prevailing 3D solutions, our approach exhibits the potential for extension to scene-level applications. This indicates that the commonsense priors from the videos significantly facilitate 3D creation. We have open-sourced all resources including the dataset, code, technical framework, and model weights: https://dropletx.github.io/.

cs.CV↗

arXiv2025

A Nyström Method for Scattering by a Two-layered Medium with a Rough Boundary

This paper is concerned with problems of scattering of time-harmonic acoustic waves by a two-layered medium with a non-locally perturbed boundary (called a rough boundary in this paper) in two dimensions, where a Dirichlet or impedance boundary condition is imposed on the boundary. The two-layered medium is composed of two unbounded media with different physical properties and the interface between the two media is considered to be a planar surface. We formulate the scattering problems considered as boundary value problems and prove the result of the well-posedness of each boundary value problem by utilizing the integral equation method associated with the two-layered Green function. Moreover, we develop a Nyström method for numerically solving the boundary value problems considered, based on the proposed integral equation formulations. We establish the convergence results of the Nyström method with the convergence rates depending on the smoothness of the rough boundary. It is worth noting that in establishing the well-posedness of the boundary value problems as well as the convergence results of the Nyström method, an essential role is played by the investigation of the asymptotic properties of the two-layered Green function for small and large arguments. Finally, numerical experiments are carried out to show the effectiveness of the Nyström method.

math.NA↗

arXiv2025

Spin light-emitting devices in a 2D magnet

Emerging two-dimensional (2D) magnetic semiconductors represent transformative platforms to explore magneto-optics and opto-spintronic applications. Though 2D opto-spintronics has attracted tremendous research efforts in spin-dependent photodetectors and non-volatile memory components, the realization of one core application - spin-modulated light-emitting device (spin-LED) - remains elusive so far. Here we successfully realize prototype spin-LED integrated with a 2D semiconducting magnet CrSBr, demonstrating considerable electroluminescence (EL) down to bilayers. Intriguingly, the EL of the spin-LED is discovered to be directly manipulated by spin-flip and spin-canting transitions. Notably, spin-flip transitions enable unprecedented hysteretic behaviors of EL characteristics, while spin-canting transitions induce EL continuous modulation with robust anisotropy. This versatile manipulation is originated from the synergy of magnetic-order mediated excitonic transitions and spintronic transport. The prototype demonstration of spin-LED establishes an indispensable scheme of opto-spintronic devices leveraging 2D spin transitions and strong excitonic effects, presenting a critical step towards integrated 2D opto-spintronics.

physics.app-ph↗

arXiv2025

Livatar-1: Real-Time Talking Heads Generation with Tailored Flow Matching

We present Livatar, a real-time audio-driven talking heads videos generation framework. Existing baselines suffer from limited lip-sync accuracy and long-term pose drift. We address these limitations with a flow matching based framework. Coupled with system optimizations, Livatar achieves competitive lip-sync quality with a 8.50 LipSync Confidence on the HDTF dataset, and reaches a throughput of 141 FPS with an end-to-end latency of 0.17s on a single A10 GPU. This makes high-fidelity avatars accessible to broader applications. Our project is available at https://www.hedra.com/ with with examples at https://h-liu1997.github.io/Livatar-1/

cs.CV↗

arXiv2025

Video Motion Graphs

We present Video Motion Graphs, a system designed to generate realistic human motion videos. Using a reference video and conditional signals such as music or motion tags, the system synthesizes new videos by first retrieving video clips with gestures matching the conditions and then generating interpolation frames to seamlessly connect clip boundaries. The core of our approach is HMInterp, a robust Video Frame Interpolation (VFI) model that enables seamless interpolation of discontinuous frames, even for complex motion scenarios like dancing. HMInterp i) employs a dual-branch interpolation approach, combining a Motion Diffusion Model for human skeleton motion interpolation with a diffusion-based video frame interpolation model for final frame generation. ii) adopts condition progressive training to effectively leverage identity strong and weak conditions, such as images and pose. These designs ensure both high video texture quality and accurate motion trajectory. Results show that our Video Motion Graphs outperforms existing generative- and retrieval-based methods for multi-modal conditioned human motion video generation. Project page can be found at https://h-liu1997.github.io/Video-Motion-Graphs/

cs.CV↗

arXiv2025

GWQ: Gradient-Aware Weight Quantization for Large Language Models

Large language models (LLMs) show impressive performance in solving complex language tasks. However, its large number of parameters presents significant challenges for the deployment. So, compressing LLMs to low bits can enable to deploy on resource-constrained devices. To address this problem, we propose gradient-aware weight quantization (GWQ), the first quantization approach for low-bit weight quantization that leverages gradients to localize outliers, requiring only a minimal amount of calibration data for outlier detection. GWQ retains the top 1\% outliers preferentially at FP16 precision, while the remaining non-outlier weights are stored in a low-bit. We widely evaluate GWQ on different task include language modeling, grounding detection, massive multitask language understanding and vision-language question and answering. Results show that models quantified by GWQ performs better than other quantization method. During quantization process, GWQ only need one calibration set to realize effective quant. Also, GWQ achieves 1.2x inference speedup in comparison to the original model and effectively reduces the inference memory.

cs.LG↗

arXiv2025

Synchronized and Fine-Grained Head for Skeleton-Based Ambiguous Action Recognition

Skeleton-based action recognition using GCNs has achieved remarkable performance, but recognizing ambiguous actions, such as "waving" and "saluting", remains a significant challenge. Existing methods typically rely on a serial combination of GCNs and TCNs, where spatial and temporal features are extracted independently, leading to an unbalanced spatial-temporal information, which hinders accurate action recognition. Moreover, existing methods for ambiguous actions often overemphasize local details, resulting in the loss of crucial global context, which further complicates the task of differentiating ambiguous actions. To address these challenges, we propose a lightweight plug-and-play module called SF-Head, inserted between GCN and TCN layers. SF-Head first conducts SSTE with a Feature Redundancy Loss (F-RL), ensuring a balanced interaction. It then performs AC-FA, with a Feature Consistency Loss (F-CL), which aligns the aggregated feature with their original spatial-temporal feature. Experimental results on NTU RGB+D 60, NTU RGB+D 120, NW-UCLA and PKU-MMD I datasets demonstrate significant improvements in distinguishing ambiguous actions.

cs.CV↗

arXiv2024

Free-viewpoint Human Animation with Pose-correlated Reference Selection

Diffusion-based human animation aims to animate a human character based on a source human image as well as driving signals such as a sequence of poses. Leveraging the generative capacity of diffusion model, existing approaches are able to generate high-fidelity poses, but struggle with significant viewpoint changes, especially in zoom-in/zoom-out scenarios where camera-character distance varies. This limits the applications such as cinematic shot type plan or camera control. We propose a pose-correlated reference selection diffusion network, supporting substantial viewpoint variations in human animation. Our key idea is to enable the network to utilize multiple reference images as input, since significant viewpoint changes often lead to missing appearance details on the human body. To eliminate the computational cost, we first introduce a novel pose correlation module to compute similarities between non-aligned target and source poses, and then propose an adaptive reference selection strategy, utilizing the attention map to identify key regions for animation generation. To train our model, we curated a large dataset from public TED talks featuring varied shots of the same character, helping the model learn synthesis for different perspectives. Our experimental results show that with the same number of reference images, our model performs favorably compared to the current SOTA methods under large viewpoint change. We further show that the adaptive reference selection is able to choose the most relevant reference regions to generate humans under free viewpoints.

cs.CV↗

arXiv2024

TANGO: Co-Speech Gesture Video Reenactment with Hierarchical Audio Motion Embedding and Diffusion Interpolation

We present TANGO, a framework for generating co-speech body-gesture videos. Given a few-minute, single-speaker reference video and target speech audio, TANGO produces high-fidelity videos with synchronized body gestures. TANGO builds on Gesture Video Reenactment (GVR), which splits and retrieves video clips using a directed graph structure - representing video frames as nodes and valid transitions as edges. We address two key limitations of GVR: audio-motion misalignment and visual artifacts in GAN-generated transition frames. In particular, (i) we propose retrieving gestures using latent feature distance to improve cross-modal alignment. To ensure the latent features could effectively model the relationship between speech audio and gesture motion, we implement a hierarchical joint embedding space (AuMoCLIP); (ii) we introduce the diffusion-based model to generate high-quality transition frames. Our diffusion model, Appearance Consistent Interpolation (ACInterp), is built upon AnimateAnyone and includes a reference motion module and homography background flow to preserve appearance consistency between generated and reference videos. By integrating these components into the graph-based retrieval framework, TANGO reliably produces realistic, audio-synchronized videos and outperforms all existing generative and retrieval methods. Our codes and pretrained models are available: \url{https://pantomatrix.github.io/TANGO/}

cs.CV↗

arXiv2024

Optimizing Cooperative path-finding: A Scalable Multi-Agent RRT* with Dynamic Potential Fields

Cooperative path-finding in multi-agent systems demands scalable solutions to navigate agents from their origins to destinations without conflict. Despite the breadth of research, scalability remains hampered by increased computational demands in complex environments. This study introduces the multi-agent RRT* potential field (MA-RRT*PF), an innovative algorithm that addresses computational efficiency and path-finding efficacy in dense scenarios. MA-RRT*PF integrates a dynamic potential field with a heuristic method, advancing obstacle avoidance and optimizing the expansion of random trees in congested spaces. The empirical evaluations highlight MA-RRT*PF's significant superiority over conventional multi-agent RRT* (MA-RRT*) in dense environments, offering enhanced performance and solution quality without compromising integrity. This work not only contributes a novel approach to the field of cooperative multi-agent path-finding but also offers a new perspective for practical applications in densely populated settings where traditional methods are less effective.

cs.MA↗

arXiv2024

EMAGE: Towards Unified Holistic Co-Speech Gesture Generation via Expressive Masked Audio Gesture Modeling

We propose EMAGE, a framework to generate full-body human gestures from audio and masked gestures, encompassing facial, local body, hands, and global movements. To achieve this, we first introduce BEAT2 (BEAT-SMPLX-FLAME), a new mesh-level holistic co-speech dataset. BEAT2 combines a MoShed SMPL-X body with FLAME head parameters and further refines the modeling of head, neck, and finger movements, offering a community-standardized, high-quality 3D motion captured dataset. EMAGE leverages masked body gesture priors during training to boost inference performance. It involves a Masked Audio Gesture Transformer, facilitating joint training on audio-to-gesture generation and masked gesture reconstruction to effectively encode audio and body gesture hints. Encoded body hints from masked gestures are then separately employed to generate facial and body movements. Moreover, EMAGE adaptively merges speech features from the audio's rhythm and content and utilizes four compositional VQ-VAEs to enhance the results' fidelity and diversity. Experiments demonstrate that EMAGE generates holistic gestures with state-of-the-art performance and is flexible in accepting predefined spatial-temporal gesture inputs, generating complete, audio-synchronized results. Our code and dataset are available https://pantomatrix.github.io/EMAGE/

cs.CV↗

arXiv2024

Prompt Learning for Multi-Label Code Smell Detection: A Promising Approach

Code smells indicate the potential problems of software quality so that developers can identify refactoring opportunities by detecting code smells. State-of-the-art approaches leverage heuristics, machine learning, and deep learning to detect code smells. However, existing approaches have not fully explored the potential of large language models (LLMs). In this paper, we propose \textit{PromptSmell}, a novel approach based on prompt learning for detecting multi-label code smell. Firstly, code snippets are acquired by traversing abstract syntax trees. Combined code snippets with natural language prompts and mask tokens, \textit{PromptSmell} constructs the input of LLMs. Secondly, to detect multi-label code smell, we leverage a label combination approach by converting a multi-label problem into a multi-classification problem. A customized answer space is added to the word list of pre-trained language models, and the probability distribution of intermediate answers is obtained by predicting the words at the mask positions. Finally, the intermediate answers are mapped to the target class labels by a verbalizer as the final classification result. We evaluate the effectiveness of \textit{PromptSmell} by answering six research questions. The experimental results demonstrate that \textit{PromptSmell} obtains an improvement of 11.17\% in $precision_{w}$ and 7.4\% in $F1_{w}$ compared to existing approaches.

cs.SE↗

arXiv2024

Global-Aware Enhanced Spatial-Temporal Graph Recurrent Networks: A New Framework For Traffic Flow Prediction

Traffic flow prediction plays a crucial role in alleviating traffic congestion and enhancing transport efficiency. While combining graph convolution networks with recurrent neural networks for spatial-temporal modeling is a common strategy in this realm, the restricted structure of recurrent neural networks limits their ability to capture global information. For spatial modeling, many prior studies learn a graph structure that is assumed to be fixed and uniform at all time steps, which may not be true. This paper introduces a novel traffic prediction framework, Global-Aware Enhanced Spatial-Temporal Graph Recurrent Network (GA-STGRN), comprising two core components: a spatial-temporal graph recurrent neural network and a global awareness layer. Within this framework, three innovative prediction models are formulated. A sequence-aware graph neural network is proposed and integrated into the Gated Recurrent Unit (GRU) to learn non-fixed graphs at different time steps and capture local temporal relationships. To enhance the model's global perception, three distinct global spatial-temporal transformer-like architectures (GST^2) are devised for the global awareness layer. We conduct extensive experiments on four real traffic datasets and the results demonstrate the superiority of our framework and the three concrete models.

cs.LG↗

arXiv2023

Equal Incremental Cost-Based Optimization Method to Enhance Efficiency for IPOP-Type Converters

Systematic optimization over a wide power range is often achieved through the combination of modules of different power levels. This paper addresses the issue of enhancing the efficiency of a multiple module system connected in parallel during operation and proposes an algorithm based on equal incremental cost for dynamic load allocation. Initially, a polynomial fitting technique is employed to fit efficiency test points for individual modules. Subsequently, the equal incremental cost-based optimization is utilized to formulate an efficiency optimization and allocation scheme for the multi-module system. A simulated annealing algorithm is applied to determine the optimal power output strategy for each module at given total power flow requirement. Finally, a dual active bridge (DAB) experimental prototype with two input-parallel-output-parallel (IPOP) configurations is constructed to validate the effectiveness of the proposed strategy. Experimental results demonstrate that under the 800W operating condition, the approach in this paper achieves an efficiency improvement of over 0.74\% by comparison with equal power sharing between both modules.

eess.SY↗

arXiv2023

Multi-Scale Spatial-Temporal Recurrent Networks for Traffic Flow Prediction

Traffic flow prediction is one of the most fundamental tasks of intelligent transportation systems. The complex and dynamic spatial-temporal dependencies make the traffic flow prediction quite challenging. Although existing spatial-temporal graph neural networks hold prominent, they often encounter challenges such as (1) ignoring the fixed graph that limits the predictive performance of the model, (2) insufficiently capturing complex spatial-temporal dependencies simultaneously, and (3) lacking attention to spatial-temporal information at different time lengths. In this paper, we propose a Multi-Scale Spatial-Temporal Recurrent Network for traffic flow prediction, namely MSSTRN, which consists of two different recurrent neural networks: the single-step gate recurrent unit and the multi-step gate recurrent unit to fully capture the complex spatial-temporal information in the traffic data under different time windows. Moreover, we propose a spatial-temporal synchronous attention mechanism that integrates adaptive position graph convolutions into the self-attention mechanism to achieve synchronous capture of spatial-temporal dependencies. We conducted extensive experiments on four real traffic datasets and demonstrated that our model achieves the best prediction accuracy with non-trivial margins compared to all the twenty baseline methods.

cs.LG↗

arXiv2023

Exploring the Mutual Influence between Self-Supervised Single-Frame and Multi-Frame Depth Estimation

Although both self-supervised single-frame and multi-frame depth estimation methods only require unlabeled monocular videos for training, the information they leverage varies because single-frame methods mainly rely on appearance-based features while multi-frame methods focus on geometric cues. Considering the complementary information of single-frame and multi-frame methods, some works attempt to leverage single-frame depth to improve multi-frame depth. However, these methods can neither exploit the difference between single-frame depth and multi-frame depth to improve multi-frame depth nor leverage multi-frame depth to optimize single-frame depth models. To fully utilize the mutual influence between single-frame and multi-frame methods, we propose a novel self-supervised training framework. Specifically, we first introduce a pixel-wise adaptive depth sampling module guided by single-frame depth to train the multi-frame model. Then, we leverage the minimum reprojection based distillation loss to transfer the knowledge from the multi-frame depth network to the single-frame network to improve single-frame depth. Finally, we regard the improved single-frame depth as a prior to further boost the performance of multi-frame depth estimation. Experimental results on the KITTI and Cityscapes datasets show that our method outperforms existing approaches in the self-supervised monocular setting.

cs.CV↗

arXiv2023

Vulnerable Smart Contract Function Locating Based on Multi-Relational Nested Graph Convolutional Network

The immutable and trustable characteristics of blockchain enable smart contracts to be applied in various fields. Unfortunately, smart contracts are subject to various vulnerabilities, which are frequently exploited by attackers, causing financial damage to users.In this paper, we study the problem of vulnerable smart contract function locating. We construct a novel Multi-Relational Nested contract Graph (MRNG) to better characterize the rich syntactic and semantic information in the smart contract code, including the relationships between data and instructions. An MRNG represents a smart contract, where each node represents a function in the smart contract and each edge describes the calling relationship between the functions. In addition, we create a Multi-Relational Function Graph (MRFG) for each function, which characterizes the corresponding function code. That is, each function is characterized as an MRFG, which corresponds to a node in the MRNG. Each MRFG uses different types of edges to represent the different control and data relationships between nodes within a function. We also propose a Multi-Relational Nested Graph Convolutional Network (MRN-GCN) to process the MRNG. MRN-GCN first extracts and aggregates features from each MRFG, using the edge-enhanced graph convolution network and self-attention mechanism. The extracted feature vector is then assigned to the corresponding node in the MRNG to obtain a new Featured Contract Graph (FCG) for the smart contract. Graph convolution is used to further extract features from the FCG. Finally, a feed forward network with a Sigmoid function is used to locate the vulnerable functions. Experimental results on the real-world smart contract datasets show that model MRN-GCN can effectively improve the accuracy, precision, recall and F1-score performance of vulnerable smart contract function locating.

cs.CR↗