Xuhang Chen

Omni-FSRNet: Omni-directional feature search and refinement network for BRAFV600E mutation prediction in thyroid ultrasound

Zhixin Liang, Junkai Zou, M.-M. Zheng, Guoheng Huang et autres

cn, mo (code pays fourni par la source)

Over-squashing as Transport Congestion: A Sandpile Dynamics Perspective

Yang Shi, Lixian Chen, Jingchao Wang, Mei Guo et autres

Message-passing graph neural networks (MP-GNNs) are widely used for learning on relational data. However, their performance drops on tasks requiring long-range interactions due to over-squash­ing, where exponential information compression overwhelms fixed-width embeddings. While existing analyses often attribute this to geometric bottlenecks under …

cn, hk (code pays fourni par la source)

Fre-QNet: Quaternion Progressive Perception Mechanism with Frequency-Guided Prompt for Blind Image Quality Assessment

Hanyu Shi, S X Li, Guoheng Huang, Yisen Zheng et autres

Blind Image Quality Assessment faces challenges in enabling computational models to mimic the hierarchical progressive perception mechanisms of the Human Visual System (HVS). Existing methods often neglect the two-stage process of HVS—global distortion identification followed by local quality evaluation—and its distinct sensitivity …

cn, mo (code pays fourni par la source)

ATRIE: Adaptive Tuning for Robust Inference and Emotion in Persona-Driven Speech Synthesis

Aoduo Li, Haoran Lv, Hongjian Xu, S Li et autres

High-fidelity character voice synthesis is a cornerstone of immersive multimedia applications, particularly for interacting with anime avatars and digital humans. However, existing systems struggle to maintain consistent persona traits across diverse emotional contexts. To bridge this gap, we present ATRIE, a unified …

cn, mo (code pays fourni par la source)

A Training-Free Paradigm for Data-Scarce Maritime Scene Classification Using Vision-Language Models

Jiabao Wu, Yujie Chen, Wentao Chen, Y. M. Lai et autres

Maritime Domain Awareness (MDA) relies heavily on data acquired from high-resolution optical spaceborne sensors; however, processing this massive quantity of sensor data via traditional supervised deep learning is severely bottlenecked by its dependency on exhaustively annotated datasets. Under extreme data scarcity, conventional …

cn, gb (code pays fourni par la source)

Beyond Shadows: A Large-Scale Benchmark and Multi-Stage Framework for High-Fidelity Facial Shadow Removal

Tailong Luo, Yihang Dong, Jiesong Bai, Jinyang Huang et autres

Facial shadows often degrade image quality and the performance of vision algorithms. Existing methods struggle to remove shadows while preserving texture, especially under complex lighting conditions, and they lack real-world paired datasets for training. We present the Augmented Shadow Face in the …

cn (code pays fourni par la source)

WAQNIQA: Wavelet-Augmented Quaternion Network for No-Reference Image Quality Assessment

Yejing Huo, Guoheng Huang, Zhiwen Yu, Xiaochen Yuan et autres

No-reference image quality assessment (NR-IQA) plays a pivotal role in computer vision by enabling image quality evaluation without reference images. While recent CNN and Transformer-based methods have advanced feature extraction, they face significant limitations. CNNs exhibit local feature bias, limiting their ability …

cn, mo (code pays fourni par la source)

Generative AI Empowers Estimation of Brain Structure-Function Connectivity for Alzheimer’s Disease

Junren Pan, Hongjie Jiang, Yanyan Shen, Xuhang Chen et autres

Fusing structural and functional brain image has become a hot topic in estimating multimodal effective connectivity for cognitive disorder identification. However, it is challenging to effectively and accurately integrate the complementary causal connectivity information using structural and functional brain images. In this …

cn, sa, jp (code pays fourni par la source)

LSFusion: Ladder-Side Attribute Composition for Multi-Aspect Controllable Text Generation

Xiaosong Yuan, Chen Shen, Shaotian Yan, Renchu Guan et autres

Multi-aspect controllable text generation (CTG), opposite to single-aspect CTG, aims to produce texts that align with multiple attributes. Incorporating attribute information through supervised fine-tuning for pre-trained language models (PLMs) on a related task is effective for single-aspect CTG. However, extending PLMs to …

cn (code pays fourni par la source)

Context Tokens are Anchors: Understanding the Repetition Curse in dMLLMs from an Information Flow Perspective

Qiyan Zhao, Xiaofeng Zhang, Shuochen Chang, Qianyu Chen et autres

Recent diffusion-based Multimodal Large Language Models (dMLLMs) suffer from high inference latency and therefore rely on caching techniques to accelerate decoding. However, the application of cache mechanisms often introduces undesirable repetitive text generation, a phenomenon we term the \textbf{Repeat Curse}. To better …

Context Tokens are Anchors: Understanding the Repetition Curse in dMLLMs from an Information Flow Perspective

Qiyan Zhao, Xiaofeng Zhang, Shuochen Chang, Qianyu Chen et autres

Recent diffusion-based Multimodal Large Language Models (dMLLMs) suffer from high inference latency and therefore rely on caching techniques to accelerate decoding. However, the application of cache mechanisms often introduces undesirable repetitive text generation, a phenomenon we term the \textbf{Repeat Curse}. To better …

ProMist-5K: A Comprehensive Dataset for Digital Emulation of Cinematic Pro-Mist Filter Effects

Yingtie Lei, Zimeng Li, Chi-Man Pun, Wangyu Wu et autres

Pro-Mist filters are widely used in cinematography for their ability to create soft halation, lower contrast, and produce a distinctive, atmospheric style. These effects are difficult to reproduce digitally due to the complex behavior of light diffusion. We present ProMist-5K, a dataset …