arXiv subjects

Zhenyu Yang

Publications and source records attributed to Zhenyu Yang.

At least 55 records · Page 3Linked to original sources

arXiv2025

DeepTelecom: A Digital-Twin Deep Learning Dataset for Channel and MIMO Applications

Domain-specific datasets are the foundation for unleashing artificial intelligence (AI)-driven wireless innovation. Yet existing wireless AI corpora are slow to produce, offer limited modeling fidelity, and cover only narrow scenario types. To address the challenges, we create DeepTelecom, a three-dimension (3D) digital-twin channel dataset. Specifically, a large language model (LLM)-assisted pipeline first builds the third level of details (LoD3) outdoor and indoor scenes with segmentable material-parameterizable surfaces. Then, DeepTelecom simulates full radio-wave propagation effects based on Sionna's ray-tracing engine. Leveraging GPU acceleration, DeepTelecom streams ray-path trajectories and real-time signal-strength heat maps, compiles them into high-frame-rate videos, and simultaneously outputs synchronized multi-view images, channel tensors, and multi-scale fading traces. By efficiently streaming large-scale, high-fidelity, and multimodal channel data, DeepTelecom not only furnishes a unified benchmark for wireless AI research but also supplies the domain-rich training substrate that enables foundation models to tightly fuse large model intelligence with future communication systems.

cs.IT↗

arXiv2025

C-MAG: Cascade Multimodal Attributed Graphs for Supply Chain Link Prediction

Workshop version accepted at KDD 2025 (AI4SupplyChain). Connecting an ever-expanding catalogue of products with suitable manufacturers and suppliers is critical for resilient, efficient global supply chains, yet traditional methods struggle to capture complex capabilities, certifications, geographic constraints, and rich multimodal data of real-world manufacturer profiles. To address these gaps, we introduce PMGraph, a public benchmark of bipartite and heterogeneous multimodal supply-chain graphs linking 8,888 manufacturers, over 70k products, more than 110k manufacturer-product edges, and over 29k product images. Building on this benchmark, we propose the Cascade Multimodal Attributed Graph C-MAG, a two-stage architecture that first aligns and aggregates textual and visual attributes into intermediate group embeddings, then propagates them through a manufacturer-product hetero-graph via multiscale message passing to enhance link prediction accuracy. C-MAG also provides practical guidelines for modality-aware fusion, preserving predictive performance in noisy, real-world settings.

cs.LG↗

arXiv2025

Efficient Agent: Optimizing Planning Capability for Multimodal Retrieval Augmented Generation

Multimodal Retrieval-Augmented Generation (mRAG) has emerged as a promising solution to address the temporal limitations of Multimodal Large Language Models (MLLMs) in real-world scenarios like news analysis and trending topics. However, existing approaches often suffer from rigid retrieval strategies and under-utilization of visual information. To bridge this gap, we propose E-Agent, an agent framework featuring two key innovations: a mRAG planner trained to dynamically orchestrate multimodal tools based on contextual reasoning, and a task executor employing tool-aware execution sequencing to implement optimized mRAG workflows. E-Agent adopts a one-time mRAG planning strategy that enables efficient information retrieval while minimizing redundant tool invocations. To rigorously assess the planning capabilities of mRAG systems, we introduce the Real-World mRAG Planning (RemPlan) benchmark. This novel benchmark contains both retrieval-dependent and retrieval-independent question types, systematically annotated with essential retrieval tools required for each instance. The benchmark's explicit mRAG planning annotations and diverse question design enhance its practical relevance by simulating real-world scenarios requiring dynamic mRAG decisions. Experiments across RemPlan and three established benchmarks demonstrate E-Agent's superiority: 13% accuracy gain over state-of-the-art mRAG methods while reducing redundant searches by 37%.

cs.AI↗

arXiv2025

X2I: Seamless Integration of Multimodal Understanding into Diffusion Transformer via Attention Distillation

Text-to-image (T2I) models are well known for their ability to produce highly realistic images, while multimodal large language models (MLLMs) are renowned for their proficiency in understanding and integrating multiple modalities. However, currently there is no straightforward and efficient framework to transfer the multimodal comprehension abilities of MLLMs to T2I models to enable them to understand multimodal inputs. In this paper, we propose the X2I framework, which endows Diffusion Transformer (DiT) models with the capability to comprehend various modalities, including multilingual text, screenshot documents, images, videos, and audio. X2I is trained using merely 100K English corpus with 160 GPU hours. Building on the DiT teacher model, we adopt an innovative distillation method to extract the inference capabilities of the teacher model and design a lightweight AlignNet structure to serve as an intermediate bridge. Compared to the teacher model, X2I shows a decrease in performance degradation of less than 1\% while gaining various multimodal understanding abilities, including multilingual to image, image to image, image-text to image, video to image, audio to image, and utilizing creative fusion to enhance imagery. Furthermore, it is applicable for LoRA training in the context of image-text to image generation, filling a void in the industry in this area. We further design a simple LightControl to enhance the fidelity of instructional image editing. Finally, extensive experiments demonstrate the effectiveness, efficiency, multifunctionality, and transferability of our X2I. The open-source code and checkpoints for X2I can be found at the following link: https://github.com/OPPO-Mente-Lab/X2I.

cs.CV↗

arXiv2025

A Dual Radiomic and Dosiomic Filtering Technique for Locoregional Radiation Pneumonitis Prediction in Breast Cancer Patients

Purpose: Radiation pneumonitis (RP) is a serious complication of intensity-modulated radiation therapy (IMRT) for breast cancer patients, underscoring the need for precise and explainable predictive models. This study presents an Explainable Dual-Omics Filtering (EDOF) model that integrates spatially localized dosiomic and radiomic features for voxel-level RP prediction. Methods: A retrospective cohort of 72 breast cancer patients treated with IMRT was analyzed, including 28 who developed RP. The EDOF model consists of two components: (1) dosiomic filtering, which extracts local dose intensity and spatial distribution features from planning dose maps, and (2) radiomic filtering, which captures texture-based features from pre-treatment CT scans. These features are jointly analyzed using the Explainable Boosting Machine (EBM), a transparent machine learning model that enables feature-specific risk evaluation. Model performance was assessed using five-fold cross-validation, reporting area under the curve (AUC), sensitivity, and specificity. Feature importance was quantified by mean absolute scores, and Partial Dependence Plots (PDPs) were used to visualize nonlinear relationships between RP risk and dual-omic features. Results: The EDOF model achieved strong predictive performance (AUC = 0.95 +- 0.01; sensitivity = 0.81 +- 0.05). The most influential features included dosiomic Intensity Mean, dosiomic Intensity Mean Absolute Deviation, and radiomic SRLGLE. PDPs revealed that RP risk increases beyond 5 Gy and rises sharply between 10-30 Gy, consistent with clinical dose thresholds. SRLGLE also captured structural heterogeneity linked to RP in specific lung regions. Conclusion: The EDOF framework enables spatially resolved, explainable RP prediction and may support personalized radiation planning to mitigate pulmonary toxicity.

physics.med-ph↗

arXiv2025

A Voxel-Wise Uncertainty-Guided Framework for Glioma Segmentation Using Spherical Projection-Based U-Net and Localized Refinement in Multi-Parametric MRI

Purpose: Accurate segmentation of glioma subregions in multi-parametric MRI (MP-MRI) is essential for diagnosis and treatment planning but remains challenging due to tumor heterogeneity and ambiguous boundaries. This study proposes an uncertainty-guided hybrid framework integrating spherical projection-based 2D modeling with targeted 3D refinement to enhance segmentation accuracy and interpretability. Methods: Using the BraTS2020 dataset (369 patients, four-modality MP-MRI), three 2D U-Nets were trained to segment enhancing tumor (ET), tumor core (TC), and whole tumor (WT). Voxel-wise uncertainty was quantified via a spherical projection-based 2D nnU-Net, capturing prediction variance across deformed inputs. A 3D sliding window was used to identify high-uncertainty regions, which were refined using a dedicated 3D nnU-Net. Final outputs combined 2D and 3D predictions through a weighted fusion optimized via Particle Swarm Optimization. Results: The proposed method outperformed standalone 2D and 3D baselines, achieving Dice scores of 0.8124 (ET), 0.7499 (TC), and 0.9055 (WT), with consistent gains in sensitivity and visual coherence. Conclusion: This work presents a novel uncertainty-aware segmentation strategy that adaptively integrates 2D and 3D modeling. By focusing refinement on ambiguous regions, it improves both efficiency and accuracy, offering broad applicability to precision neuro-oncology and other high-stakes medical imaging tasks.

physics.med-ph↗

arXiv2025

An Explainable Neural Radiomic Sequence Model with Spatiotemporal Continuity for Quantifying 4DCT-based Pulmonary Ventilation

Accurate evaluation of regional lung ventilation is essential for the management and treatment of lung cancer patients, supporting assessments of pulmonary function, optimization of therapeutic strategies, and monitoring of treatment response. Currently, ventilation scintigraphy using nuclear medicine techniques is widely employed in clinical practice; however, it is often time-consuming, costly, and entails additional radiation exposure. In this study, we propose an explainable neural radiomic sequence model to identify regions of compromised pulmonary ventilation based on four-dimensional computed tomography (4DCT). A cohort of 45 lung cancer patients from the VAMPIRE dataset was analyzed. For each patient, lung volumes were segmented from 4DCT, and voxel-wise radiomic features (56-dimensional) were extracted across the respiratory cycle to capture local intensity and texture dynamics, forming temporal radiomic sequences. Ground truth ventilation defects were delineated voxel-wise using Galligas-PET and DTPA-SPECT. To identify compromised regions, we developed a temporal saliency-enhanced explainable long short-term memory (LSTM) network trained on the radiomic sequences. Temporal saliency maps were generated to highlight key features contributing to the model's predictions. The proposed model demonstrated robust performance, achieving average (range) Dice similarity coefficients of 0.78 (0.74-0.79) for 25 PET cases and 0.78 (0.74-0.82) for 20 SPECT cases. The temporal saliency map explained three key radiomic sequences in ventilation quantification: during lung exhalation, compromised pulmonary function region typically exhibits (1) an increasing trend of intensity and (2) a decreasing trend of homogeneity, in contrast to healthy lung tissue.

physics.med-ph↗

arXiv2025

H2VU-Benchmark: A Comprehensive Benchmark for Hierarchical Holistic Video Understanding

With the rapid development of multimodal models, the demand for assessing video understanding capabilities has been steadily increasing. However, existing benchmarks for evaluating video understanding exhibit significant limitations in coverage, task diversity, and scene adaptability. These shortcomings hinder the accurate assessment of models' comprehensive video understanding capabilities. To tackle this challenge, we propose a hierarchical and holistic video understanding (H2VU) benchmark designed to evaluate both general video and online streaming video comprehension. This benchmark contributes three key features: Extended video duration: Spanning videos from brief 3-second clips to comprehensive 1.5-hour recordings, thereby bridging the temporal gaps found in current benchmarks. Comprehensive assessment tasks: Beyond traditional perceptual and reasoning tasks, we have introduced modules for countercommonsense comprehension and trajectory state tracking. These additions test the models' deep understanding capabilities beyond mere prior knowledge. Enriched video data: To keep pace with the rapid evolution of current AI agents, we have expanded first-person streaming video datasets. This expansion allows for the exploration of multimodal models' performance in understanding streaming videos from a first-person perspective. Extensive results from H2VU reveal that existing multimodal large language models (MLLMs) possess substantial potential for improvement in our newly proposed evaluation tasks. We expect that H2VU will facilitate advancements in video understanding research by offering a comprehensive and in-depth analysis of MLLMs.

cs.CV↗

arXiv2025

Embedding Radiomics into Vision Transformers for Multimodal Medical Image Classification

Background: Deep learning has significantly advanced medical image analysis, with Vision Transformers (ViTs) offering a powerful alternative to convolutional models by modeling long-range dependencies through self-attention. However, ViTs are inherently data-intensive and lack domain-specific inductive biases, limiting their applicability in medical imaging. In contrast, radiomics provides interpretable, handcrafted descriptors of tissue heterogeneity but suffers from limited scalability and integration into end-to-end learning frameworks. In this work, we propose the Radiomics-Embedded Vision Transformer (RE-ViT) that combines radiomic features with data-driven visual embeddings within a ViT backbone. Purpose: To develop a hybrid RE-ViT framework that integrates radiomics and patch-wise ViT embeddings through early fusion, enhancing robustness and performance in medical image classification. Methods: Following the standard ViT pipeline, images were divided into patches. For each patch, handcrafted radiomic features were extracted and fused with linearly projected pixel embeddings. The fused representations were normalized, positionally encoded, and passed to the ViT encoder. A learnable [CLS] token aggregated patch-level information for classification. We evaluated RE-ViT on three public datasets (including BUSI, ChestXray2017, and Retinal OCT) using accuracy, macro AUC, sensitivity, and specificity. RE-ViT was benchmarked against CNN-based (VGG-16, ResNet) and hybrid (TransMed) models. Results: RE-ViT achieved state-of-the-art results: on BUSI, AUC=0.950+/-0.011; on ChestXray2017, AUC=0.989+/-0.004; on Retinal OCT, AUC=0.986+/-0.001, which outperforms other comparison models. Conclusions: The RE-ViT framework effectively integrates radiomics with ViT architectures, demonstrating improved performance and generalizability across multimodal medical image classification tasks.

physics.med-ph↗

arXiv2025

Improved Visual-Spatial Reasoning via R1-Zero-Like Training

Increasing attention has been placed on improving the reasoning capacities of multi-modal large language models (MLLMs). As the cornerstone for AI agents that function in the physical realm, video-based visual-spatial intelligence (VSI) emerges as one of the most pivotal reasoning capabilities of MLLMs. This work conducts a first, in-depth study on improving the visual-spatial reasoning of MLLMs via R1-Zero-like training. Technically, we first identify that the visual-spatial reasoning capacities of small- to medium-sized Qwen2-VL models cannot be activated via Chain of Thought (CoT) prompts. We then incorporate GRPO training for improved visual-spatial reasoning, using the carefully curated VSI-100k dataset, following DeepSeek-R1-Zero. During the investigation, we identify the necessity to keep the KL penalty (even with a small value) in GRPO. With just 120 GPU hours, our vsGRPO-2B model, fine-tuned from Qwen2-VL-2B, can outperform the base model by 12.1% and surpass GPT-4o. Moreover, our vsGRPO-7B model, fine-tuned from Qwen2-VL-7B, achieves performance comparable to that of the best open-source model LLaVA-NeXT-Video-72B. Additionally, we compare vsGRPO to supervised fine-tuning and direct preference optimization baselines and observe strong performance superiority. The code and dataset will be available soon.

cs.CV↗

arXiv2025

Compiler Optimization Testing Based on Optimization-Guided Equivalence Transformations

Compiler optimization techniques are inherently complex, and rigorous testing of compiler optimization implementation is critical. Recent years have witnessed the emergence of testing approaches for uncovering incorrect optimization bugs, but these approaches rely heavily on the differential testing mechanism, which requires comparing outputs across multiple compilers. This dependency gives rise to important limitations, including that (1) the tested functionality must be consistently implemented across all compilers and (2) shared bugs remain undetected. Thus, false alarms can be produced and significant manual efforts will be required. To overcome the limitations, we propose a metamorphic testing approach inspired by compiler optimizations. The approach is driven by how to maximize compiler optimization opportunities while effectively judging optimization correctness. Specifically, our approach first employs tailored code construction strategies to generate input programs that satisfy optimization conditions, and then applies various compiler optimization transformations to create semantically equivalent test programs. By comparing the outputs of pre- and post-transformation programs, this approach effectively identifies incorrect optimization bugs. We conducted a preliminary evaluation of this approach on GCC and LLVM, and we have successfully detected five incorrect optimization bugs at the time of writing. This result demonstrates the effectiveness and potential of our approach.

cs.SE↗

arXiv2025

LandMarkSystem Technical Report

3D reconstruction is vital for applications in autonomous driving, virtual reality, augmented reality, and the metaverse. Recent advancements such as Neural Radiance Fields(NeRF) and 3D Gaussian Splatting (3DGS) have transformed the field, yet traditional deep learning frameworks struggle to meet the increasing demands for scene quality and scale. This paper introduces LandMarkSystem, a novel computing framework designed to enhance multi-scale scene reconstruction and rendering. By leveraging a componentized model adaptation layer, LandMarkSystem supports various NeRF and 3DGS structures while optimizing computational efficiency through distributed parallel computing and model parameter offloading. Our system addresses the limitations of existing frameworks, providing dedicated operators for complex 3D sparse computations, thus facilitating efficient training and rapid inference over extensive scenes. Key contributions include a modular architecture, a dynamic loading strategy for limited resources, and proven capabilities across multiple representative algorithms.This comprehensive solution aims to advance the efficiency and effectiveness of 3D reconstruction tasks.To facilitate further research and collaboration, the source code and documentation for the LandMarkSystem project are publicly available in an open-source repository, accessing the repository at: https://github.com/InternLandMark/LandMarkSystem.

cs.CV↗

arXiv2025

Layton: Latent Consistency Tokenizer for 1024-pixel Image Reconstruction and Generation by 256 Tokens

Image tokenization has significantly advanced visual generation and multimodal modeling, particularly when paired with autoregressive models. However, current methods face challenges in balancing efficiency and fidelity: high-resolution image reconstruction either requires an excessive number of tokens or compromises critical details through token reduction. To resolve this, we propose Latent Consistency Tokenizer (Layton) that bridges discrete visual tokens with the compact latent space of pre-trained Latent Diffusion Models (LDMs), enabling efficient representation of 1024x1024 images using only 256 tokens-a 16 times compression over VQGAN. Layton integrates a transformer encoder, a quantized codebook, and a latent consistency decoder. Direct application of LDM as the decoder results in color and brightness discrepancies. Thus, we convert it to latent consistency decoder, reducing multi-step sampling to 1-2 steps for direct pixel-level supervision. Experiments demonstrate Layton's superiority in high-fidelity reconstruction, with 10.8 reconstruction Frechet Inception Distance on MSCOCO-2017 5K benchmark for 1024x1024 image reconstruction. We also extend Layton to a text-to-image generation model, LaytonGen, working in autoregression. It achieves 0.73 score on GenEval benchmark, surpassing current state-of-the-art methods. Project homepage: https://github.com/OPPO-Mente-Lab/Layton

cs.CV↗

arXiv2025

FasterCache: Training-Free Video Diffusion Model Acceleration with High Quality

In this paper, we present \textbf{\textit{FasterCache}}, a novel training-free strategy designed to accelerate the inference of video diffusion models with high-quality generation. By analyzing existing cache-based methods, we observe that \textit{directly reusing adjacent-step features degrades video quality due to the loss of subtle variations}. We further perform a pioneering investigation of the acceleration potential of classifier-free guidance (CFG) and reveal significant redundancy between conditional and unconditional features within the same timestep. Capitalizing on these observations, we introduce FasterCache to substantially accelerate diffusion-based video generation. Our key contributions include a dynamic feature reuse strategy that preserves both feature distinction and temporal continuity, and CFG-Cache which optimizes the reuse of conditional and unconditional outputs to further enhance inference speed without compromising video quality. We empirically evaluate FasterCache on recent video diffusion models. Experimental results show that FasterCache can significantly accelerate video generation (\eg 1.67$\times$ speedup on Vchitect-2.0) while keeping video quality comparable to the baseline, and consistently outperform existing methods in both inference speed and video quality.

cs.CV↗

arXiv2025

Enantiomer-Selective Magnetoresistance in Chiral Gold Nanocrystals by Magnetic Control of Surface Potentials

Chiral nanomaterials offer intriguing possibilities for novel electronic and chemical applications. Here, we report the discovery of an enantiomer-selective magnetoresistance effect in chiral gold nanocrystals. Based on precise control of nanocrystal chiral morphology using amino acid-directed synthesis, we demonstrate that an external magnetic field can dramatically modulate resistance in an enantiomer-specific manner. For a given enantiomer, a magnetic field in one direction alters the resistance by dozens of times, while the opposite field direction leaves it unchanged. This asymmetric response reverses for the opposite enantiomer and are reproduced in both single nanocrystals by conduction atomic force microscopy and nanocrystal thin films in solid state devices. We attribute this phenomenon to a chirality-driven charge pumping effect, where the interplay between the chiral morphology and the magnetic field selectively modifies the surface potential. The magnitude and sign of the magnetoresistance can be further tuned by the surface chemistry of the nanocrystal, as demonstrated through sulfide treatment. Our findings reveal a new form of chirality-dependent magnetoresistance, distinct from previously known effects such as chirality-induced spin selectivity and electric magnetochiral anisotropy. The ability to remotely control surface potentials of chiral nanostructures using magnetic fields could enable novel approaches in catalysis, drug delivery, and nanoelectronics.

cond-mat.mes-hall↗

arXiv2025

Binary Neural Networks for Large Language Model: A Survey

Large language models (LLMs) have wide applications in the field of natural language processing(NLP), such as GPT-4 and Llama. However, with the exponential growth of model parameter sizes, LLMs bring significant resource overheads. Low-bit quantization, as a key technique, reduces memory usage and computational demands by decreasing the bit-width of model parameters, activations, and gradients. Previous quantization methods for LLMs have largely employed Post-Training Quantization (PTQ) and Quantization-Aware Training (QAT). PTQ does not require any retraining of the original model, while QAT involves optimizing precision during training to achieve the best quantization parameters. The BitNet team proposed a radically different approach, where quantization is performed from the start of model training, utilizing low-precision binary weights during the training process. This approach has led to the emergence of many binary quantization techniques for large language models. This paper provides a comprehensive review of these binary quantization techniques. Specifically, we will introduce binary quantization techniques in deep neural networks and further explore their application to LLMs, reviewing their various contributions, implementations, and applications.

cs.CL↗

arXiv2025

GS-Cache: A GS-Cache Inference Framework for Large-scale Gaussian Splatting Models

Rendering large-scale 3D Gaussian Splatting (3DGS) model faces significant challenges in achieving real-time, high-fidelity performance on consumer-grade devices. Fully realizing the potential of 3DGS in applications such as virtual reality (VR) requires addressing critical system-level challenges to support real-time, immersive experiences. We propose GS-Cache, an end-to-end framework that seamlessly integrates 3DGS's advanced representation with a highly optimized rendering system. GS-Cache introduces a cache-centric pipeline to eliminate redundant computations, an efficiency-aware scheduler for elastic multi-GPU rendering, and optimized CUDA kernels to overcome computational bottlenecks. This synergy between 3DGS and system design enables GS-Cache to achieve up to 5.35x performance improvement, 35% latency reduction, and 42% lower GPU memory usage, supporting 2K binocular rendering at over 120 FPS with high visual quality. By bridging the gap between 3DGS's representation power and the demands of VR systems, GS-Cache establishes a scalable and efficient framework for real-time neural rendering in immersive environments.

cs.CV↗

arXiv2025

GlyphDraw2: Automatic Generation of Complex Glyph Posters with Diffusion Models and Large Language Models

Posters play a crucial role in marketing and advertising by enhancing visual communication and brand visibility, making significant contributions to industrial design. With the latest advancements in controllable T2I diffusion models, increasing research has focused on rendering text within synthesized images. Despite improvements in text rendering accuracy, the field of automatic poster generation remains underexplored. In this paper, we propose an automatic poster generation framework with text rendering capabilities leveraging LLMs, utilizing a triple-cross attention mechanism based on alignment learning. This framework aims to create precise poster text within a detailed contextual background. Additionally, the framework supports controllable fonts, adjustable image resolution, and the rendering of posters with descriptions and text in both English and Chinese.Furthermore, we introduce a high-resolution font dataset and a poster dataset with resolutions exceeding 1024 pixels. Our approach leverages the SDXL architecture. Extensive experiments validate our method's capability in generating poster images with complex and contextually rich backgrounds.Codes is available at https://github.com/OPPO-Mente-Lab/GlyphDraw2.

cs.CV↗