arXiv subjects

Zhe Zhang

Publications and source records attributed to Zhe Zhang.

At least 91 records · Page 5Linked to original sources

arXiv2025

Unified Domain Adaptive Semantic Segmentation

Unsupervised Domain Adaptive Semantic Segmentation (UDA-SS) aims to transfer the supervision from a labeled source domain to an unlabeled target domain. The majority of existing UDA-SS works typically consider images whilst recent attempts have extended further to tackle videos by modeling the temporal dimension. Although the two lines of research share the major challenges -- overcoming the underlying domain distribution shift, their studies are largely independent, resulting in fragmented insights, a lack of holistic understanding, and missed opportunities for cross-pollination of ideas. This fragmentation prevents the unification of methods, leading to redundant efforts and suboptimal knowledge transfer across image and video domains. Under this observation, we advocate unifying the study of UDA-SS across video and image scenarios, enabling a more comprehensive understanding, synergistic advancements, and efficient knowledge sharing. To that end, we explore the unified UDA-SS from a general data augmentation perspective, serving as a unifying conceptual framework, enabling improved generalization, and potential for cross-pollination of ideas, ultimately contributing to the overall progress and practical impact of this field of research. Specifically, we propose a Quad-directional Mixup (QuadMix) method, characterized by tackling distinct point attributes and feature inconsistencies through four-directional paths for intra- and inter-domain mixing in a feature space. To deal with temporal shifts with videos, we incorporate optical flow-guided feature aggregation across spatial and temporal dimensions for fine-grained domain alignment. Extensive experiments show that our method outperforms the state-of-the-art works by large margins on four challenging UDA-SS benchmarks. Our source code and models will be released at https://github.com/ZHE-SAPI/UDASS.

cs.CV↗

arXiv2025

MMRQA: Signal-Enhanced Multimodal Large Language Models for MRI Quality Assessment

Magnetic resonance imaging (MRI) quality assessment is crucial for clinical decision-making, yet remains challenging due to data scarcity and protocol variability. Traditional approaches face fundamental trade-offs: signal-based methods like MRIQC provide quantitative metrics but lack semantic understanding, while deep learning approaches achieve high accuracy but sacrifice interpretability. To address these limitations, we introduce the Multimodal MRI Quality Assessment (MMRQA) framework, pioneering the integration of multimodal large language models (MLLMs) with acquisition-aware signal processing. MMRQA combines three key innovations: robust metric extraction via MRQy augmented with simulated artifacts, structured transformation of metrics into question-answer pairs using Qwen, and parameter-efficient fusion through Low-Rank Adaptation (LoRA) of LLaVA-OneVision. Evaluated on MR-ART, FastMRI, and MyConnectome benchmarks, MMRQA achieves state-of-the-art performance with strong zero-shot generalization, as validated by comprehensive ablation studies. By bridging quantitative analysis with semantic reasoning, our framework generates clinically interpretable outputs that enhance quality control in dynamic medical settings.

cs.CV↗

arXiv2025

Wan-Animate: Unified Character Animation and Replacement with Holistic Replication

We introduce Wan-Animate, a unified framework for character animation and replacement. Given a character image and a reference video, Wan-Animate can animate the character by precisely replicating the expressions and movements of the character in the video to generate high-fidelity character videos. Alternatively, it can integrate the animated character into the reference video to replace the original character, replicating the scene's lighting and color tone to achieve seamless environmental integration. Wan-Animate is built upon the Wan model. To adapt it for character animation tasks, we employ a modified input paradigm to differentiate between reference conditions and regions for generation. This design unifies multiple tasks into a common symbolic representation. We use spatially-aligned skeleton signals to replicate body motion and implicit facial features extracted from source images to reenact expressions, enabling the generation of character videos with high controllability and expressiveness. Furthermore, to enhance environmental integration during character replacement, we develop an auxiliary Relighting LoRA. This module preserves the character's appearance consistency while applying the appropriate environmental lighting and color tone. Experimental results demonstrate that Wan-Animate achieves state-of-the-art performance. We are committed to open-sourcing the model weights and its source code.

cs.CV↗

arXiv2025

HD-MoE: Hybrid and Dynamic Parallelism for Mixture-of-Expert LLMs with 3D Near-Memory Processing

Large Language Models (LLMs) with Mixture-of-Expert (MoE) architectures achieve superior model performance with reduced computation costs, but at the cost of high memory capacity and bandwidth requirements. Near-Memory Processing (NMP) accelerators that stack memory directly on the compute through hybrid bonding have demonstrated high bandwidth with high energy efficiency, becoming a promising architecture for MoE models. However, as NMP accelerators comprise distributed memory and computation, how to map the MoE computation directly determines the LLM inference efficiency. Existing parallel mapping strategies, including Tensor Parallelism (TP) and Expert Parallelism (EP), suffer from either high communication costs or unbalanced computation utilization, leading to inferior efficiency. The dynamic routing mechanism of MoE LLMs further aggravates the efficiency challenges. Therefore, in this paper, we propose HD-MoE to automatically optimize the MoE parallel computation across an NMP accelerator. HD-MoE features an offline automatic hybrid parallel mapping algorithm and an online dynamic scheduling strategy to reduce the communication costs while maximizing the computation utilization. With extensive experimental results, we demonstrate that HD-MoE achieves a speedup ranging from 1.1x to 1.8x over TP, 1.1x to 1.5x over EP, and 1.0x to 1.4x over the baseline Hybrid TP-EP with Compute-Balanced parallelism strategies.

cs.PF↗

arXiv2025

${C}^{3}$-GS: Learning Context-aware, Cross-dimension, Cross-scale Feature for Generalizable Gaussian Splatting

Generalizable Gaussian Splatting aims to synthesize novel views for unseen scenes without per-scene optimization. In particular, recent advancements utilize feed-forward networks to predict per-pixel Gaussian parameters, enabling high-quality synthesis from sparse input views. However, existing approaches fall short in encoding discriminative, multi-view consistent features for Gaussian predictions, which struggle to construct accurate geometry with sparse views. To address this, we propose $\mathbf{C}^{3}$-GS, a framework that enhances feature learning by incorporating context-aware, cross-dimension, and cross-scale constraints. Our architecture integrates three lightweight modules into a unified rendering pipeline, improving feature fusion and enabling photorealistic synthesis without requiring additional supervision. Extensive experiments on benchmark datasets validate that $\mathbf{C}^{3}$-GS achieves state-of-the-art rendering quality and generalization ability. Code is available at: https://github.com/YuhsiHu/C3-GS.

cs.CV↗

arXiv2025

Wan-S2V: Audio-Driven Cinematic Video Generation

Current state-of-the-art (SOTA) methods for audio-driven character animation demonstrate promising performance for scenarios primarily involving speech and singing. However, they often fall short in more complex film and television productions, which demand sophisticated elements such as nuanced character interactions, realistic body movements, and dynamic camera work. To address this long-standing challenge of achieving film-level character animation, we propose an audio-driven model, which we refere to as Wan-S2V, built upon Wan. Our model achieves significantly enhanced expressiveness and fidelity in cinematic contexts compared to existing approaches. We conducted extensive experiments, benchmarking our method against cutting-edge models such as Hunyuan-Avatar and Omnihuman. The experimental results consistently demonstrate that our approach significantly outperforms these existing solutions. Additionally, we explore the versatility of our method through its applications in long-form video generation and precise video lip-sync editing.

cs.CV↗

arXiv2025

ShortListing Model: A Streamlined SimplexDiffusion for Discrete Variable Generation

Generative modeling of discrete variables is challenging yet crucial for applications in natural language processing and biological sequence design. We introduce the Shortlisting Model (SLM), a novel simplex-based diffusion model inspired by progressive candidate pruning. SLM operates on simplex centroids, reducing generation complexity and enhancing scalability. Additionally, SLM incorporates a flexible implementation of classifier-free guidance, enhancing unconditional generation performance. Extensive experiments on DNA promoter and enhancer design, protein design, character-level and large-vocabulary language modeling demonstrate the competitive performance and strong potential of SLM. Our code can be found at https://github.com/GenSI-THUAIR/SLM

cs.LG↗

arXiv2025

WEC-DG: Multi-Exposure Wavelet Correction Method Guided by Degradation Description

Multi-exposure correction technology is essential for restoring images affected by insufficient or excessive lighting, enhancing the visual experience by improving brightness, contrast, and detail richness. However, current multi-exposure correction methods often encounter challenges in addressing intra-class variability caused by diverse lighting conditions, shooting environments, and weather factors, particularly when processing images captured at a single exposure level. To enhance the adaptability of these models under complex imaging conditions, this paper proposes a Wavelet-based Exposure Correction method with Degradation Guidance (WEC-DG). Specifically, we introduce a degradation descriptor within the Exposure Consistency Alignment Module (ECAM) at both ends of the processing pipeline to ensure exposure consistency and achieve final alignment. This mechanism effectively addresses miscorrected exposure anomalies caused by existing methods' failure to recognize 'blurred' exposure degradation. Additionally, we investigate the light-detail decoupling properties of the wavelet transform to design the Exposure Restoration and Detail Reconstruction Module (EDRM), which processes low-frequency information related to exposure enhancement before utilizing high-frequency information as a prior guide for reconstructing spatial domain details. This serial processing strategy guarantees precise light correction and enhances detail recovery. Extensive experiments conducted on multiple public datasets demonstrate that the proposed method outperforms existing algorithms, achieving significant performance improvements and validating its effectiveness and practical applicability.

cs.CV↗

arXiv2025

A Cross-Perspective Annotated Dataset for Dynamic Object-Level Attention Modeling in Cloud Gaming

Cloud gaming has gained popularity as it provides high-quality gaming experiences on thin hardware, such as phones and tablets. Transmitting gameplay frames at high resolutions and ultra-low latency is the key to guaranteeing players' quality of experience (QoE). Numerous studies have explored deep learning (DL) techniques to address this challenge. The efficiency of these DL-based approaches is highly affected by the dataset. However, existing datasets usually focus on the positions of objects while ignoring semantic relationships with other objects and their unique features. In this paper, we present a game dataset by collecting gameplay clips from Grand Theft Auto (GTA) V, and annotating the player's interested objects during the gameplay. Based on the collected data, we analyze several factors that have an impact on player's interest and identify that the player's in-game speed, object's size, and object's speed are the main factors. The dataset is available at https://drive.google.com/drive/folders/1idH251a2K-hGGd3pKjX-3Gx5o_rUqLC4?usp=sharing

cs.DB↗

arXiv2025

CIVQLLIE: Causal Intervention with Vector Quantization for Low-Light Image Enhancement

Images captured in nighttime scenes suffer from severely reduced visibility, hindering effective content perception. Current low-light image enhancement (LLIE) methods face significant challenges: data-driven end-to-end mapping networks lack interpretability or rely on unreliable prior guidance, struggling under extremely dark conditions, while physics-based methods depend on simplified assumptions that often fail in complex real-world scenarios. To address these limitations, we propose CIVQLLIE, a novel framework that leverages the power of discrete representation learning through causal reasoning. We achieve this through Vector Quantization (VQ), which maps continuous image features to a discrete codebook of visual tokens learned from large-scale high-quality images. This codebook serves as a reliable prior, encoding standardized brightness and color patterns that are independent of degradation. However, direct application of VQ to low-light images fails due to distribution shifts between degraded inputs and the learned codebook. Therefore, we propose a multi-level causal intervention approach to systematically correct these shifts. First, during encoding, our Pixel-level Causal Intervention (PCI) module intervenes to align low-level features with the brightness and color distributions expected by the codebook. Second, a Feature-aware Causal Intervention (FCI) mechanism with Low-frequency Selective Attention Gating (LSAG) identifies and enhances channels most affected by illumination degradation, facilitating accurate codebook token matching while enhancing the encoder's generalization performance through flexible feature-level intervention. Finally, during decoding, the High-frequency Detail Reconstruction Module (HDRM) leverages structural information preserved in the matched codebook representations to reconstruct fine details using deformable convolution techniques.

cs.CV↗

arXiv2025

Neutrino-jet correlations in charged-current SIDIS

Charged-current deep inelastic scattering plays a significant role in determining parton distribution functions with flavour separation.In this work, we present a systematic calculation of the charged-current semi-inclusive deep inelastic scattering (SIDIS) in the $eN$ collinear frame up to twist-3 level at leading order. Semi-inclusive refers to the process in which a jet is detected in addition to the scattered neutrino. We focus on neutrino-jet correlations in our calculation. We first present the differential cross section in terms of structure functions, followed by the differential cross section expressed in term of transverse momentum dependent parton distribution functions. We derive the complete set of azimuthal asymmetries and intrinsic asymmetries. We also introduce an observable $A^C$, defined as the ratio of the difference to the sum of differential cross sections for electron and positron semi-inclusive deep inelastic scattering.We notice that $A^C$ provides a sensitive probe for valence and sea quark distribution functions

hep-ph↗

arXiv2025

Predictive Hydrodynamic Simulations for Laser Direct-drive Implosion Experiments via Artificial Intelligence

This work presents predictive hydrodynamic simulations empowered by artificial intelligence (AI) for laser driven implosion experiments, taking the double-cone ignition (DCI) scheme as an example. A Transformer-based deep learning model MULTI-Net is established to predict implosion features according to laser waveforms and target radius. A Physics-Informed Decoder (PID) is proposed for high-dimensional sampling, significantly reducing the prediction errors compared to Latin hypercube sampling. Applied to DCI experiments conducted on the SG-II Upgrade facility, the MULTI-Net model is able to predict the implosion dynamics measured by the x-ray streak camera. It is found that an effective laser absorption factor about 65\% is suitable for the one-dimensional simulations of the DCI-R10 experiments. For shot 33, the mean implosion velocity and collided plasma density reached 195 km/s and 117 g/cc, respectively. This study demonstrates a data-driven AI framework that enhances the prediction ability of simulations for complicated laser fusion experiments.

physics.plasm-ph↗

arXiv2025

CWNet: Causal Wavelet Network for Low-Light Image Enhancement

Traditional Low-Light Image Enhancement (LLIE) methods primarily focus on uniform brightness adjustment, often neglecting instance-level semantic information and the inherent characteristics of different features. To address these limitations, we propose CWNet (Causal Wavelet Network), a novel architecture that leverages wavelet transforms for causal reasoning. Specifically, our approach comprises two key components: 1) Inspired by the concept of intervention in causality, we adopt a causal reasoning perspective to reveal the underlying causal relationships in low-light enhancement. From a global perspective, we employ a metric learning strategy to ensure causal embeddings adhere to causal principles, separating them from non-causal confounding factors while focusing on the invariance of causal factors. At the local level, we introduce an instance-level CLIP semantic loss to precisely maintain causal factor consistency. 2) Based on our causal analysis, we present a wavelet transform-based backbone network that effectively optimizes the recovery of frequency information, ensuring precise enhancement tailored to the specific attributes of wavelet transforms. Extensive experiments demonstrate that CWNet significantly outperforms current state-of-the-art methods across multiple datasets, showcasing its robust performance across diverse scenes. Code is available at https://github.com/bywlzts/CWNet-Causal-Wavelet-Network.

cs.CV↗

arXiv2025

MIDI-VALLE: Improving Expressive Piano Performance Synthesis Through Neural Codec Language Modelling

Generating expressive audio performances from music scores requires models to capture both instrument acoustics and human interpretation. Traditional music performance synthesis pipelines follow a two-stage approach, first generating expressive performance MIDI from a score, then synthesising the MIDI into audio. However, the synthesis models often struggle to generalise across diverse MIDI sources, musical styles, and recording environments. To address these challenges, we propose MIDI-VALLE, a neural codec language model adapted from the VALLE framework, which was originally designed for zero-shot personalised text-to-speech (TTS) synthesis. For performance MIDI-to-audio synthesis, we improve the architecture to condition on a reference audio performance and its corresponding MIDI. Unlike previous TTS-based systems that rely on piano rolls, MIDI-VALLE encodes both MIDI and audio as discrete tokens, facilitating a more consistent and robust modelling of piano performances. Furthermore, the model's generalisation ability is enhanced by training on an extensive and diverse piano performance dataset. Evaluation results show that MIDI-VALLE significantly outperforms a state-of-the-art baseline, achieving over 75% lower Frechet Audio Distance on the ATEPP and Maestro datasets. In the listening test, MIDI-VALLE received 202 votes compared to 58 for the baseline, demonstrating improved synthesis quality and generalisation across diverse performance MIDI inputs.

cs.SD↗

arXiv2025

VLM-TDP: VLM-guided Trajectory-conditioned Diffusion Policy for Robust Long-Horizon Manipulation

Diffusion policy has demonstrated promising performance in the field of robotic manipulation. However, its effectiveness has been primarily limited in short-horizon tasks, and its performance significantly degrades in the presence of image noise. To address these limitations, we propose a VLM-guided trajectory-conditioned diffusion policy (VLM-TDP) for robust and long-horizon manipulation. Specifically, the proposed method leverages state-of-the-art vision-language models (VLMs) to decompose long-horizon tasks into concise, manageable sub-tasks, while also innovatively generating voxel-based trajectories for each sub-task. The generated trajectories serve as a crucial conditioning factor, effectively steering the diffusion policy and substantially enhancing its performance. The proposed Trajectory-conditioned Diffusion Policy (TDP) is trained on trajectories derived from demonstration data and validated using the trajectories generated by the VLM. Simulation experimental results indicate that our method significantly outperforms classical diffusion policies, achieving an average 44% increase in success rate, over 100% improvement in long-horizon tasks, and a 20% reduction in performance degradation in challenging conditions, such as noisy images or altered environments. These findings are further reinforced by our real-world experiments, where the performance gap becomes even more pronounced in long-horizon tasks. Videos are available on https://youtu.be/g0T6h32OSC8

cs.RO↗

arXiv2025

MAPS: Advancing Multi-Modal Reasoning in Expert-Level Physical Science

Pre-trained on extensive text and image corpora, current Multi-Modal Large Language Models (MLLM) have shown strong capabilities in general visual reasoning tasks. However, their performance is still lacking in physical domains that require understanding diagrams with complex physical structures and quantitative analysis based on multi-modal information. To address this, we develop a new framework, named Multi-Modal Scientific Reasoning with Physics Perception and Simulation (MAPS) based on an MLLM. MAPS decomposes expert-level multi-modal reasoning task into physical diagram understanding via a Physical Perception Model (PPM) and reasoning with physical knowledge via a simulator. The PPM module is obtained by fine-tuning a visual language model using carefully designed synthetic data with paired physical diagrams and corresponding simulation language descriptions. At the inference stage, MAPS integrates the simulation language description of the input diagram provided by PPM and results obtained through a Chain-of-Simulation process with MLLM to derive the underlying rationale and the final answer. Validated using our collected college-level circuit analysis problems, MAPS significantly improves reasoning accuracy of MLLM and outperforms all existing models. The results confirm MAPS offers a promising direction for enhancing multi-modal scientific reasoning ability of MLLMs. We will release our code, model and dataset used for our experiments upon publishing of this paper.

cs.AI↗

arXiv2025

Mitigating Language Mismatch in SSL-Based Speaker Anonymization

Speaker anonymization aims to protect speaker identity while preserving content information and the intelligibility of speech. However, most speaker anonymization systems (SASs) are developed and evaluated using only English, resulting in degraded utility for other languages. This paper investigates language mismatch in SASs for Japanese and Mandarin speech. First, we fine-tune a self-supervised learning (SSL)-based content encoder with Japanese speech to verify effective language adaptation. Then, we propose fine-tuning a multilingual SSL model with Japanese speech and evaluating the SAS in Japanese and Mandarin. Downstream experiments show that fine-tuning an English-only SSL model with the target language enhances intelligibility while maintaining privacy and that multilingual SSL further extends SASs' utility across different languages. These findings highlight the importance of language adaptation and multilingual pre-training of SSLs for robust multilingual speaker anonymization.

eess.AS↗

arXiv2025

ReF-LLE: Personalized Low-Light Enhancement via Reference-Guided Deep Reinforcement Learning

Low-light image enhancement presents two primary challenges: 1) Significant variations in low-light images across different conditions, and 2) Enhancement levels influenced by subjective preferences and user intent. To address these issues, we propose ReF-LLE, a novel personalized low-light image enhancement method that operates in the Fourier frequency domain and incorporates deep reinforcement learning. ReF-LLE is the first to integrate deep reinforcement learning into this domain. During training, a zero-reference image evaluation strategy is introduced to score enhanced images, providing reward signals that guide the model to handle varying degrees of low-light conditions effectively. In the inference phase, ReF-LLE employs a personalized adaptive iterative strategy, guided by the zero-frequency component in the Fourier domain, which represents the overall illumination level. This strategy enables the model to adaptively adjust low-light images to align with the illumination distribution of a user-provided reference image, ensuring personalized enhancement results. Extensive experiments on benchmark datasets demonstrate that ReF-LLE outperforms state-of-the-art methods, achieving superior perceptual quality and adaptability in personalized low-light image enhancement.

cs.CV↗