arXiv subjects

Zhe Chen

Publications and source records attributed to Zhe Chen.

At least 127 records · Page 7Linked to original sources

arXiv2025

Enhancing the Reasoning Ability of Multimodal Large Language Models via Mixed Preference Optimization

Existing open-source multimodal large language models (MLLMs) generally follow a training process involving pre-training and supervised fine-tuning. However, these models suffer from distribution shifts, which limit their multimodal reasoning, particularly in the Chain-of-Thought (CoT) performance. To address this, we introduce a preference optimization (PO) process to enhance the multimodal reasoning capabilities of MLLMs. Specifically, (1) on the data side, we design an automated preference data construction pipeline to create MMPR, a high-quality, large-scale multimodal reasoning preference dataset; and (2) on the model side, we explore integrating PO with MLLMs, developing a simple yet effective method, termed Mixed Preference Optimization (MPO), which boosts multimodal CoT performance. Our approach enhances the multimodal reasoning abilities of both InternVL2-8B and InternVL2-76B. Notably, our model, InternVL2-8B-MPO, achieves an accuracy of 67.0 on MathVista, outperforming InternVL2-8B by 8.7 points and achieving performance comparable to the 10$\times$ larger InternVL2-76B. We hope this study could inspire further advancements in MLLMs. Code, data, and model are released.

cs.CL↗

arXiv2025

Vision-RWKV: Efficient and Scalable Visual Perception with RWKV-Like Architectures

Transformers have revolutionized computer vision and natural language processing, but their high computational complexity limits their application in high-resolution image processing and long-context analysis. This paper introduces Vision-RWKV (VRWKV), a model adapted from the RWKV model used in the NLP field with necessary modifications for vision tasks. Similar to the Vision Transformer (ViT), our model is designed to efficiently handle sparse inputs and demonstrate robust global processing capabilities, while also scaling up effectively, accommodating both large-scale parameters and extensive datasets. Its distinctive advantage lies in its reduced spatial aggregation complexity, which renders it exceptionally adept at processing high-resolution images seamlessly, eliminating the necessity for windowing operations. Our evaluations demonstrate that VRWKV surpasses ViT's performance in image classification and has significantly faster speeds and lower memory usage processing high-resolution inputs. In dense prediction tasks, it outperforms window-based models, maintaining comparable speeds. These results highlight VRWKV's potential as a more efficient alternative for visual perception tasks. Code is released at https://github.com/OpenGVLab/Vision-RWKV.

cs.CV↗

arXiv2025

WHU-Synthetic: A Synthetic Perception Dataset for 3-D Multitask Model Research

End-to-end models capable of handling multiple sub-tasks in parallel have become a new trend, thereby presenting significant challenges and opportunities for the integration of multiple tasks within the domain of 3D vision. The limitations of 3D data acquisition conditions have not only restricted the exploration of many innovative research problems but have also caused existing 3D datasets to predominantly focus on single tasks. This has resulted in a lack of systematic approaches and theoretical frameworks for 3D multi-task learning, with most efforts merely serving as auxiliary support to the primary task. In this paper, we introduce WHU-Synthetic, a large-scale 3D synthetic perception dataset designed for multi-task learning, from the initial data augmentation (upsampling and depth completion), through scene understanding (segmentation), to macro-level tasks (place recognition and 3D reconstruction). Collected in the same environmental domain, we ensure inherent alignment across sub-tasks to construct multi-task models without separate training methods. Besides, we implement several novel settings, making it possible to realize certain ideas that are difficult to achieve in real-world scenarios. This supports more adaptive and robust multi-task perception tasks, such as sampling on city-level models, providing point clouds with different densities, and simulating temporal changes. Using our dataset, we conduct several experiments to investigate mutual benefits between sub-tasks, revealing new observations, challenges, and opportunities for future research. The dataset is accessible at https://github.com/WHU-USI3DV/WHU-Synthetic.

cs.CV↗

arXiv2025

GMAI-VL & GMAI-VL-5.5M: A Large Vision-Language Model and A Comprehensive Multimodal Dataset Towards General Medical AI

Despite significant advancements in general AI, its effectiveness in the medical domain is limited by the lack of specialized medical knowledge. To address this, we formulate GMAI-VL-5.5M, a multimodal medical dataset created by converting hundreds of specialized medical datasets with various annotations into high-quality image-text pairs. This dataset offers comprehensive task coverage, diverse modalities, and rich image-text data. Building upon this dataset, we develop GMAI-VL, a general medical vision-language model, with a three-stage training strategy that enhances the integration of visual and textual information. This approach significantly improves the model's ability to process multimodal data, supporting accurate diagnoses and clinical decision-making. Experiments show that GMAI-VL achieves state-of-the-art performance across various multimodal medical tasks, including visual question answering and medical image diagnosis.

cs.CV↗

arXiv2025

Fùxì: A Benchmark for Evaluating Language Models on Ancient Chinese Text Understanding and Generation

Ancient Chinese text processing presents unique challenges for large language models (LLMs) due to its distinct linguistic features, complex structural constraints, and rich cultural context. While existing benchmarks have primarily focused on evaluating comprehension through multiple-choice questions, there remains a critical gap in assessing models' generative capabilities in classical Chinese. We introduce Fùxì, a comprehensive benchmark that evaluates both understanding and generation capabilities across 21 diverse tasks. Our benchmark distinguishes itself through three key contributions: (1) balanced coverage of both comprehension and generation tasks, including novel tasks like poetry composition and couplet completion, (2) specialized evaluation metrics designed specifically for classical Chinese text generation, combining rule-based verification with fine-tuned LLM evaluators, and (3) a systematic assessment framework that considers both linguistic accuracy and cultural authenticity. Through extensive evaluation of state-of-the-art LLMs, we reveal significant performance gaps between understanding and generation tasks, with models achieving promising results in comprehension but struggling considerably in generation tasks, particularly those requiring deep cultural knowledge and adherence to classical formats. Our findings highlight the current limitations in ancient Chinese text processing and provide insights for future model development. The benchmark, evaluation toolkit, and baseline results are publicly available to facilitate research in this domain.

cs.CL↗

arXiv2025

VisualPRM: An Effective Process Reward Model for Multimodal Reasoning

We introduce VisualPRM, an advanced multimodal Process Reward Model (PRM) with 8B parameters, which improves the reasoning abilities of existing Multimodal Large Language Models (MLLMs) across different model scales and families with Best-of-N (BoN) evaluation strategies. Specifically, our model improves the reasoning performance of three types of MLLMs and four different model scales. Even when applied to the highly capable InternVL2.5-78B, it achieves a 5.9-point improvement across seven multimodal reasoning benchmarks. Experimental results show that our model exhibits superior performance compared to Outcome Reward Models and Self-Consistency during BoN evaluation. To facilitate the training of multimodal PRMs, we construct a multimodal process supervision dataset VisualPRM400K using an automated data pipeline. For the evaluation of multimodal PRMs, we propose VisualProcessBench, a benchmark with human-annotated step-wise correctness labels, to measure the abilities of PRMs to detect erroneous steps in multimodal reasoning tasks. We hope that our work can inspire more future research and contribute to the development of MLLMs. Our model, data, and benchmark are released in https://internvl.github.io/blog/2025-03-13-VisualPRM/.

cs.CV↗

arXiv2025

DSVD: Dynamic Self-Verify Decoding for Faithful Generation in Large Language Models

The reliability of large language models remains a critical challenge, particularly due to their susceptibility to hallucinations and factual inaccuracies during text generation. Existing solutions either underutilize models' self-correction with preemptive strategies or use costly post-hoc verification. To further explore the potential of real-time self-verification and correction, we present Dynamic Self-Verify Decoding (DSVD), a novel decoding framework that enhances generation reliability through real-time hallucination detection and efficient error correction. DSVD integrates two key components: (1) parallel self-verification architecture for continuous quality assessment, (2) dynamic rollback mechanism for targeted error recovery. Extensive experiments across five benchmarks demonstrate DSVD's effectiveness, achieving significant improvement in truthfulness (Quesetion-Answering) and factual accuracy (FActScore). Results show the DSVD can be further incorporated with existing faithful decoding methods to achieve stronger performance. Our work establishes that real-time self-verification during generation offers a viable path toward more trustworthy language models without sacrificing practical deployability.

cs.CL↗

arXiv2025

Patterning Silver Nanowire Network via the Gibbs-Thomson Effect

As transparent electrodes, patterned silver nanowire (AgNW) networks suffer from noticeable pattern visibility, which is an unsettled issue for practical applications such as display. Here, we introduce a Gibbs-Thomson effect (GTE)-based patterning method to effectively reduce pattern visibility. Unlike conventional top-down and bottom-up strategies that rely on selective etching, removal, or deposition of AgNWs, our approach focuses on fragmenting nanowires primarily at the junctions through the GTE. This is realized by modifying AgNWs with a compound of diphenyliodonium nitrate and silver nitrate, which aggregates into nanoparticles at the junctions of AgNWs. These nanoparticles can boost the fragmentation of nanowires at the junctions under an ultralow temperature (75°C), allow pattern transfer through a photolithographic masking operation, and enhance plasmonic welding during UV exposure. The resultant patterned electrodes have trivial differences in transmittance (ΔT = 1.4%) and haze (ΔH = 0.3%) between conductive and insulative regions, with high-resolution patterning size down to 10 μm. To demonstrate the practicality of this novel method, we constructed a highly transparent, optoelectrical interactive tactile e-skin using the patterned AgNW electrodes.

eess.SY↗

arXiv2025

RL-OGM-Parking: Lidar OGM-Based Hybrid Reinforcement Learning Planner for Autonomous Parking

Autonomous parking has become a critical application in automatic driving research and development. Parking operations often suffer from limited space and complex environments, requiring accurate perception and precise maneuvering. Traditional rule-based parking algorithms struggle to adapt to diverse and unpredictable conditions, while learning-based algorithms lack consistent and stable performance in various scenarios. Therefore, a hybrid approach is necessary that combines the stability of rule-based methods and the generalizability of learning-based methods. Recently, reinforcement learning (RL) based policy has shown robust capability in planning tasks. However, the simulation-to-reality (sim-to-real) transfer gap seriously blocks the real-world deployment. To address these problems, we employ a hybrid policy, consisting of a rule-based Reeds-Shepp (RS) planner and a learning-based reinforcement learning (RL) planner. A real-time LiDAR-based Occupancy Grid Map (OGM) representation is adopted to bridge the sim-to-real gap, leading the hybrid policy can be applied to real-world systems seamlessly. We conducted extensive experiments both in the simulation environment and real-world scenarios, and the result demonstrates that the proposed method outperforms pure rule-based and learning-based methods. The real-world experiment further validates the feasibility and efficiency of the proposed method.

cs.RO↗

arXiv2025

Large Memory Network for Recommendation

Modeling user behavior sequences in recommender systems is essential for understanding user preferences over time, enabling personalized and accurate recommendations for improving user retention and enhancing business values. Despite its significance, there are two challenges for current sequential modeling approaches. From the spatial dimension, it is difficult to mutually perceive similar users' interests for a generalized intention understanding; from the temporal dimension, current methods are generally prone to forgetting long-term interests due to the fixed-length input sequence. In this paper, we present Large Memory Network (LMN), providing a novel idea by compressing and storing user history behavior information in a large-scale memory block. With the elaborated online deployment strategy, the memory block can be easily scaled up to million-scale in the industry. Extensive offline comparison experiments, memory scaling up experiments, and online A/B test on Douyin E-Commerce Search (ECS) are performed, validating the superior performance of LMN. Currently, LMN has been fully deployed in Douyin ECS, serving millions of users each day.

cs.IR↗

arXiv2025

Incomplete Modality Disentangled Representation for Ophthalmic Disease Grading and Diagnosis

Ophthalmologists typically require multimodal data sources to improve diagnostic accuracy in clinical decisions. However, due to medical device shortages, low-quality data and data privacy concerns, missing data modalities are common in real-world scenarios. Existing deep learning methods tend to address it by learning an implicit latent subspace representation for different modality combinations. We identify two significant limitations of these methods: (1) implicit representation constraints that hinder the model's ability to capture modality-specific information and (2) modality heterogeneity, causing distribution gaps and redundancy in feature representations. To address these, we propose an Incomplete Modality Disentangled Representation (IMDR) strategy, which disentangles features into explicit independent modal-common and modal-specific features by guidance of mutual information, distilling informative knowledge and enabling it to reconstruct valuable missing semantics and produce robust multimodal representations. Furthermore, we introduce a joint proxy learning module that assists IMDR in eliminating intra-modality redundancy by exploiting the extracted proxies from each class. Experiments on four ophthalmology multimodal datasets demonstrate that the proposed IMDR outperforms the state-of-the-art methods significantly.

cs.CV↗

arXiv2025

An Experimental Study of Noise Reduction in Wind Turbine Airfoils with Serrated Trailing Edges

This study explores the noise reduction achieved by airfoils with serrated trailing edges in a low turbulence wind tunnel, focusing on acoustic spectral characteristics and wake flow field measurements. We analyze the effects of various factors, including Reynolds number, angle of attack, serration parameters, and model type, on sound power levels and far-field radiation patterns. Our findings reveal that serrated trailing edges significantly reduce blunt vortex shedding noise and laminar separation bubble noise across a broader frequency range, particularly in the mid-to-high frequency range, with reductions delineated by two boundaries. Interestingly, the serration geometry exhibits minimal impact on noise reduction, which varies with the angle of attack and airfoil profile across all tested conditions. Additionally, while serrations effectively lower noise levels, especially at higher frequencies, they do not significantly alter the airfoil's acoustic directivity patterns. Measurements of wake flow velocity spectra demonstrate a clear correlation between reduced wake turbulence and noise reduction, as serrated edges decrease the power spectral density of turbulent velocity fluctuations, effectively disrupting larger vortex structures responsible for noise generation. These valuable insights contribute to understanding the acoustic benefits of serrated trailing edges.

physics.flu-dyn↗

arXiv2025

HoVLE: Unleashing the Power of Monolithic Vision-Language Models with Holistic Vision-Language Embedding

The rapid advance of Large Language Models (LLMs) has catalyzed the development of Vision-Language Models (VLMs). Monolithic VLMs, which avoid modality-specific encoders, offer a promising alternative to the compositional ones but face the challenge of inferior performance. Most existing monolithic VLMs require tuning pre-trained LLMs to acquire vision abilities, which may degrade their language capabilities. To address this dilemma, this paper presents a novel high-performance monolithic VLM named HoVLE. We note that LLMs have been shown capable of interpreting images, when image embeddings are aligned with text embeddings. The challenge for current monolithic VLMs actually lies in the lack of a holistic embedding module for both vision and language inputs. Therefore, HoVLE introduces a holistic embedding module that converts visual and textual inputs into a shared space, allowing LLMs to process images in the same way as texts. Furthermore, a multi-stage training strategy is carefully designed to empower the holistic embedding module. It is first trained to distill visual features from a pre-trained vision encoder and text embeddings from the LLM, enabling large-scale training with unpaired random images and text tokens. The whole model further undergoes next-token prediction on multi-modal data to align the embeddings. Finally, an instruction-tuning stage is incorporated. Our experiments show that HoVLE achieves performance close to leading compositional models on various benchmarks, outperforming previous monolithic models by a large margin. Model available at https://huggingface.co/OpenGVLab/HoVLE.

cs.CV↗

arXiv2025

RotateKV: Accurate and Robust 2-Bit KV Cache Quantization for LLMs via Outlier-Aware Adaptive Rotations

Key-Value (KV) cache facilitates efficient large language models (LLMs) inference by avoiding recomputation of past KVs. As the batch size and context length increase, the oversized KV caches become a significant memory bottleneck, highlighting the need for efficient compression. Existing KV quantization rely on fine-grained quantization or the retention of a significant portion of high bit-widths caches, both of which compromise compression ratio and often fail to maintain robustness at extremely low average bit-widths. In this work, we explore the potential of rotation technique for 2-bit KV quantization and propose RotateKV, which achieves accurate and robust performance through the following innovations: (i) Outlier-Aware Rotation, which utilizes channel-reordering to adapt the rotations to varying channel-wise outlier distributions without sacrificing the computational efficiency of the fast Walsh-Hadamard transform (FWHT); (ii) Pre-RoPE Grouped-Head Rotation, which mitigates the impact of rotary position embedding (RoPE) on proposed outlier-aware rotation and further smooths outliers across heads; (iii) Attention-Sink-Aware Quantization, which leverages the massive activations to precisely identify and protect attention sinks. RotateKV achieves less than 0.3 perplexity (PPL) degradation with 2-bit quantization on WikiText-2 using LLaMA-2-13B, maintains strong CoT reasoning and long-context capabilities, with less than 1.7\% degradation on GSM8K, outperforming existing methods even at lower average bit-widths. RotateKV also showcases a 3.97x reduction in peak memory usage, supports 5.75x larger batch sizes, and achieves a 2.32x speedup in decoding stage.

cs.LG↗

arXiv2025

AKVQ-VL: Attention-Aware KV Cache Adaptive 2-Bit Quantization for Vision-Language Models

Vision-language models (VLMs) show remarkable performance in multimodal tasks. However, excessively long multimodal inputs lead to oversized Key-Value (KV) caches, resulting in significant memory consumption and I/O bottlenecks. Previous KV quantization methods for Large Language Models (LLMs) may alleviate these issues but overlook the attention saliency differences of multimodal tokens, resulting in suboptimal performance. In this paper, we investigate the attention-aware token saliency patterns in VLM and propose AKVQ-VL. AKVQ-VL leverages the proposed Text-Salient Attention (TSA) and Pivot-Token-Salient Attention (PSA) patterns to adaptively allocate bit budgets. Moreover, achieving extremely low-bit quantization requires effectively addressing outliers in KV tensors. AKVQ-VL utilizes the Walsh-Hadamard transform (WHT) to construct outlier-free KV caches, thereby reducing quantization difficulty. Evaluations of 2-bit quantization on 12 long-context and multimodal tasks demonstrate that AKVQ-VL maintains or even improves accuracy, outperforming LLM-oriented methods. AKVQ-VL can reduce peak memory usage by 2.13x, support up to 3.25x larger batch sizes and 2.46x throughput.

cs.CL↗

arXiv2025

FLORA: Formal Language Model Enables Robust Training-free Zero-shot Object Referring Analysis

Object Referring Analysis (ORA), commonly known as referring expression comprehension, requires the identification and localization of specific objects in an image based on natural descriptions. Unlike generic object detection, ORA requires both accurate language understanding and precise visual localization, making it inherently more complex. Although recent pre-trained large visual grounding detectors have achieved significant progress, they heavily rely on extensively labeled data and time-consuming learning. To address these, we introduce a novel, training-free framework for zero-shot ORA, termed FLORA (Formal Language for Object Referring and Analysis). FLORA harnesses the inherent reasoning capabilities of large language models (LLMs) and integrates a formal language model - a logical framework that regulates language within structured, rule-based descriptions - to provide effective zero-shot ORA. More specifically, our formal language model (FLM) enables an effective, logic-driven interpretation of object descriptions without necessitating any training processes. Built upon FLM-regulated LLM outputs, we further devise a Bayesian inference framework and employ appropriate off-the-shelf interpretive models to finalize the reasoning, delivering favorable robustness against LLM hallucinations and compelling ORA performance in a training-free manner. In practice, our FLORA boosts the zero-shot performance of existing pretrained grounding detectors by up to around 45%. Our comprehensive evaluation across different challenging datasets also confirms that FLORA consistently surpasses current state-of-the-art zero-shot methods in both detection and segmentation tasks associated with zero-shot ORA. We believe our probabilistic parsing and reasoning of the LLM outputs elevate the reliability and interpretability of zero-shot ORA. We shall release codes upon publication.

cs.CV↗

arXiv2025

Swin-X2S: Reconstructing 3D Shape from 2D Biplanar X-ray with Swin Transformers

The conversion from 2D X-ray to 3D shape holds significant potential for improving diagnostic efficiency and safety. However, existing reconstruction methods often rely on hand-crafted features, manual intervention, and prior knowledge, resulting in unstable shape errors and additional processing costs. In this paper, we introduce Swin-X2S, an end-to-end deep learning method for directly reconstructing 3D segmentation and labeling from 2D biplanar orthogonal X-ray images. Swin-X2S employs an encoder-decoder architecture: the encoder leverages 2D Swin Transformer for X-ray information extraction, while the decoder employs 3D convolution with cross-attention to integrate structural features from orthogonal views. A dimension-expanding module is introduced to bridge the encoder and decoder, ensuring a smooth conversion from 2D pixels to 3D voxels. We evaluate proposed method through extensive qualitative and quantitative experiments across nine publicly available datasets covering four anatomies (femur, hip, spine, and rib), with a total of 54 categories. Significant improvements over previous methods have been observed not only in the segmentation and labeling metrics but also in the clinically relevant parameters that are of primary concern in practical applications, which demonstrates the promise of Swin-X2S to provide an effective option for anatomical shape reconstruction in clinical scenarios. Code implementation is available at: \url{https://github.com/liukuan5625/Swin-X2S}.

cs.CV↗

arXiv2025

Constructing 4D Radio Map in LEO Satellite Networks with Limited Samples

Recently, Low Earth Orbit (LEO) satellite networks (i.e., non-terrestrial network (NTN)), such as Starlink, have been successfully deployed to provide broader coverage than terrestrial networks (TN). Due to limited spectrum resources, TN and NTN may soon share the same spectrum. Therefore, fine-grained spectrum monitoring is crucial for spectrum sharing and interference avoidance. To this end, constructing a 4D radio map (RM) including three spatial dimensions and signal spectra is important. However, this requires the large deployment of sensors, and high-speed analog-to-digital converters for extensive spatial signal collection and wide power spectrum acquisition, respectively. To address these challenges, we propose a deep unsupervised learning framework without ground truths labeling requirement, DeepRM, comprised of neural compressive sensing (CS) and tensor decomposition (TD) algorithms. Firstly, we map the CS process into the optimization of a neural networksassociated loss function, and design a sparsity-performance balance training algorithm to reconstruct a wide power spectrum under limited sub-Nquist samples. Secondly, according to the output of neural CS algorithm, we also utilize neural networks to perform TD, and construct the 3D RM for each frequency, even under very sparse sensor deployment. Extensive evaluations show that DeepRM achieves lower error than its corresponding state-of-the-art baselines, especially with limited samples.

cs.NI↗