Juncheng Li

Diffusion-based Laplacian frequency-aware network for low-light image enhancement

Li Zhou, Wenjie Li, Juncheng Li, G.F. Gao et autres

cn, tw (code pays fourni par la source)

Dispersible reduced graphene oxide enhances the biodegradation of sulfamethoxazole by white rot fungus Phanerochaete chrysosporium

Yunxuan Hu, Juncheng Li, Xiting Wang, Lejie Pan et autres

cn (code pays fourni par la source)

Unified multi-prototype network with pretrained swin transformer for visual and audio open set recognition

Haiyan Yang, Sheng Li, Juncheng Li, Jun Shi et autres

cn, jp (code pays fourni par la source)

Survey on Deep Face Restoration: From Non-blind to Blind and Beyond

Wenjie Li, Mei Wang, Kai Zhang, Juncheng Li et autres

Face restoration (FR) is a specialized field within image restoration that aims to recover low-quality (LQ) face images into high-quality (HQ) face images. Recent advances in deep learning technology have led to significant progress in FR methods. In this article, we begin …

cn, sg (code pays fourni par la source)

Evolve the Method, Not the Prompts: Evolutionary Synthesis of Jailbreak Attacks on LLMs

Xin Wang, Juncheng Li, Yixu Wang, Jie Li et autres

Automated red teaming frameworks for Large Language Models (LLMs) have become increasingly sophisticated, yet many still formulate attack optimization primarily in the prompt space. In other words, these methods mainly search for better attack wording or better strategy choices, but they do …

WEAVE: Unleashing and Benchmarking the In-context Interleaved Comprehension and Generation

Wei Chow, Jiachun Pan, Yongyuan Liang, Mingze Zhou et autres

Recent advances in unified multimodal models (UMMs) have enabled impressive progress in visual comprehension and generation. However, existing datasets and benchmarks focus primarily on single-turn interactions, failing to capture the multi-turn, context-dependent nature of real-world image creation and editing. To address this …

Uncertainty‐Aware Graph Self‐Training for Autism Spectrum Disorder Classification in Multiple Centers

Qianhui Yang, Jun Wang, Jiale Dun, Juncheng Li et autres

ABSTRACT Classical self‐training methods for graph convolutional networks (GCNs) assume that both labeled and unlabeled data follow the identical distribution. However, these works do not work well when they are used in medical applications such as classifying autism spectrum disorder (ASD) in …

cn (code pays fourni par la source)

Cross Paradigm Representation and Alignment Transformer for Image Deraining

Juncheng Li, Guangwei Gao, Guo-Jun Qi

Transformer-based networks have achieved strong performance in low-level vision tasks like image deraining by utilizing spatial or channel-wise self-attention. However, irregular rain patterns and complex geometric overlaps challenge single-paradigm architectures, necessitating a unified framework to integrate complementary global-local and spatial-channel representations. To …

cn (code pays fourni par la source)

Iris: Breaking GUI Complexity with Adaptive Focus and Self-Refining

Zhiqi Ge, Juncheng Li, Xiaoli Pang, Minghe Gao et autres

Digital agents are increasingly employed to automate tasks in interactive digital environments such as web pages, software applications, and operating systems. While text-based agents built on Large Language Models (LLMs) often require frequent updates due to platform-specific APIs, visual agents leveraging Multimodal …

cn, sg (code pays fourni par la source)

Mastering Collaborative Multi-Modal Data Selection: A Focus on Informativeness, Uniqueness, and Representativeness

Qifan Yu, Zhongqi Quentin Yue, Yang Wu, Wenqiao Zhang et autres

Instruction tuning fine-tunes pre-trained Multi-modal Large Language Models (MLLMs) to handle real-world tasks. However, the rapid expansion of visual instruction datasets introduces data redundancy, leading to excessive computational costs. We propose a collaborative framework, DataTailor, which leverages three key principles--informativeness, uniqueness, and …

cn, sg (code pays fourni par la source)

W-MTHN: Wavelet-Driven Mamba-Transformer Hybrid Network for Image Deraining

Lie Shan, Mingju Chen, Chenxi Dong, Dacheng Zhou et autres

Image deraining plays a vital role in ensuring clear vision under adverse weather conditions. In this paper, we propose W-MTHN, a novel Wavelet-driven Mamba-Transformer Hybrid Network, designed for robust and efficient image deraining. Our key contribution lies in the synergistic integration of …

cn (code pays fourni par la source)

Fast Thinking for Large Language Models

Haoyu Zheng, Zhuonan Wang, Yuqian Yuan, Wenqiao Zhang et autres

Reasoning-oriented Large Language Models (LLMs) often rely on generating explicit tokens step by step, and their effectiveness typically hinges on large-scale supervised fine-tuning or reinforcement learning. While Chain-of-Thought (CoT) techniques substantially enhance performance on complex reasoning tasks, they remain inefficient, requiring long …