Author name cluster

Zedong Wang

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

8 papers

2 author rows

AAAI Conference 2025 Conference Paper

Prior-guided Hierarchical Harmonization Network for Efficient Image Dehazing

Xiongfei Su
Siyuan Li
Yuning Cui
Miao Cao
Yulun Zhang
Zheng Chen
Zongliang Wu
Zedong Wang

Image dehazing is a crucial task that involves the enhancement of degraded images to recover their sharpness and textures. While vision Transformers have exhibited impressive results in diverse dehazing tasks, their quadratic complexity and lack of dehazing priors pose significant drawbacks for real-world applications. In this paper, guided by triple priors, Bright Channel Prior (BCP), Dark Channel Prior (DCP), and Histogram Equalization (HE), we propose a Prior-guided Hierarchical Harmonization Network (PGHHNet) for image dehazing. PGHNet is built upon the UNet-like architecture with an efficient encoder and decoder, consisting of two module types: (1) Prior aggregation module that injects BCP/DCP and selects diverse contexts with gating attention. (2) Feature harmonization modules that subtract low-frequency components from spatial and channel aspects and learn more informative feature distributions to equalize the feature maps. Inspired by observing the sparsity of BCP/DCP and the histogram equalization, we harmonize the deep features using a histogram equation-guided module and further leverage BCP/DCP to guide spatial attention through a sandwich module as the bottleneck. Comprehensive experiments demonstrate that our model efficiently attains the highest level of performance among existing methods across four different datasets for image dehazing tasks.

PDF Details DOI

IJCAI Conference 2024 Conference Paper

LongVQ: Long Sequence Modeling with Vector Quantization on Structured Memory

Zicheng Liu
Li Wang
Siyuan Li
Zedong Wang
Haitao Lin
Stan Z. Li

Transformer models have been successful in various sequence processing tasks, but the self-attention mechanism's computational cost limits its practicality for long sequences. Although there are existing attention variants that improve computational efficiency, they have a limited ability to abstract global information effectively based on their hand-crafted mixing strategies. On the other hand, state-space models (SSMs) are tailored for long sequences but cannot capture complicated local information. Therefore, the combination of them as a unified token mixer is a trend in recent long-sequence models. However, the linearized attention degrades performance significantly even when equipped with SSMs. To address the issue, we propose a new method called LongVQ. LongVQ uses the vector quantization (VQ) technique to compress the global abstraction as a length-fixed codebook, enabling the linear-time computation of the attention matrix. This technique effectively maintains dynamic global and local patterns, which helps to complement the lack of long-range dependency issues. Our experiments on the Long Range Arena benchmark, autoregressive language modeling, and image and speech classification demonstrate the effectiveness of LongVQ. Our model achieves significant improvements over other sequence models, including variants of Transformers, Convolutions, and recent State Space Models.

PDF Details DOI

ICLR Conference 2024 Conference Paper

By contextualizing the kernel as global as possible, Modern ConvNets have shown great potential in computer vision tasks. However, recent progress on \textit{multi-order game-theoretic interaction} within deep neural networks (DNNs) reveals the representation bottleneck of modern ConvNets, where the expressive interactions have not been effectively encoded with the increased kernel size. To tackle this challenge, we propose a new family of modern ConvNets, dubbed MogaNet, for discriminative visual representation learning in pure ConvNet-based models with favorable complexity-performance trade-offs. MogaNet encapsulates conceptually simple yet effective convolutions and gated aggregation into a compact module, where discriminative features are efficiently gathered and contextualized adaptively. MogaNet exhibits great scalability, impressive efficiency of parameters, and competitive performance compared to state-of-the-art ViTs and ConvNets on ImageNet and various downstream vision benchmarks, including COCO object detection, ADE20K semantic segmentation, 2D\&3D human pose estimation, and video prediction. Notably, MogaNet hits 80.0\% and 87.8\% accuracy with 5.2M and 181M parameters on ImageNet-1K, outperforming ParC-Net and ConvNeXt-L, while saving 59\% FLOPs and 17M parameters, respectively. The source code is available at https://github.com/Westlake-AI/MogaNet.

Details

ICLR Conference 2024 Conference Paper

SemiReward: A General Reward Model for Semi-supervised Learning

Siyuan Li 0002
Weiyang Jin
Zedong Wang
Fang Wu 0002
Zicheng Liu 0006
Cheng Tan 0012
Stan Z. Li

Semi-supervised learning (SSL) has witnessed great progress with various improvements in the self-training framework with pseudo labeling. The main challenge is how to distinguish high-quality pseudo labels against the confirmation bias. However, existing pseudo-label selection strategies are limited to pre-defined schemes or complex hand-crafted policies specially designed for classification, failing to achieve high-quality labels, fast convergence, and task versatility simultaneously. To these ends, we propose a Semi-supervised Reward framework (SemiReward) that predicts reward scores to evaluate and filter out high-quality pseudo labels, which is pluggable to mainstream SSL methods in wide task types and scenarios. To mitigate confirmation bias, SemiReward is trained online in two stages with a generator model and subsampling strategy. With classification and regression tasks on 13 standard SSL benchmarks across three modalities, extensive experiments verify that SemiReward achieves significant performance gains and faster convergence speeds upon Pseudo Label, FlexMatch, and Free/SoftMatch. Code and models are available at https://github.com/Westlake-AI/SemiReward.

Details

ICML Conference 2024 Conference Paper

Short-Long Convolutions Help Hardware-Efficient Linear Attention to Focus on Long Sequences

Zicheng Liu 0006
Siyuan Li 0002
Li Wang
Zedong Wang
Yunfan Liu 0002
Stan Z. Li

To mitigate the computational complexity in the self-attention mechanism on long sequences, linear attention utilizes computation tricks to achieve linear complexity, while state space models (SSMs) popularize a favourable practice of using non-data-dependent memory pattern, i. e. , emphasize the near and neglect the distant, to processing sequences. Recent studies have shown the priorities by combining them as one. However, the efficiency of linear attention remains only at the theoretical level in a causal setting, and SSMs require various designed constraints to operate effectively on specific data. Therefore, in order to unveil the true power of the hybrid design, the following two issues need to be addressed: (1) hardware-efficient implementation for linear attention and (2) stabilization of SSMs. To achieve this, we leverage the thought of tiling and hierarchy to propose CHELA (short-long Convolutions with Hardware-Efficient Linear Attention), which replaces SSMs with short-long convolutions and implements linear attention in a divide-and-conquer manner. This approach enjoys global abstraction and data-dependent selection from stable SSM and linear attention while maintaining real linear complexity. Our comprehensive experiments on the Long Range Arena benchmark and language modeling tasks demonstrate the effectiveness of the proposed method.

Details

ICML Conference 2024 Conference Paper

VQDNA: Unleashing the Power of Vector Quantization for Multi-Species Genomic Sequence Modeling

Siyuan Li 0002
Zedong Wang
Zicheng Liu 0006
Di Wu 0057
Cheng Tan 0012
Jiangbin Zheng 0002
Yufei Huang 0002
Stan Z. Li

Similar to natural language models, pre-trained genome language models are proposed to capture the underlying intricacies within genomes with unsupervised sequence modeling. They have become essential tools for researchers and practitioners in biology. However, the hand-crafted tokenization policies used in these models may not encode the most discriminative patterns from the limited vocabulary of genomic data. In this paper, we introduce VQDNA, a general-purpose framework that renovates genome tokenization from the perspective of genome vocabulary learning. By leveraging vector-quantized codebook as learnable vocabulary, VQDNA can adaptively tokenize genomes into pattern-aware embeddings in an end-to-end manner. To further push its limits, we propose Hierarchical Residual Quantization (HRQ), where varying scales of codebooks are designed in a hierarchy to enrich the genome vocabulary in a coarse-to-fine manner. Extensive experiments on 32 genome datasets demonstrate VQDNA’s superiority and favorable parameter efficiency compared to existing genome language models. Notably, empirical analysis of SARS-CoV-2 mutations reveals the fine-grained pattern awareness and biological significance of learned HRQ vocabulary, highlighting its untapped potential for broader applications in genomics.

Details

NeurIPS Conference 2023 Conference Paper

OpenSTL: A Comprehensive Benchmark of Spatio-Temporal Predictive Learning

Cheng Tan
Siyuan Li
Zhangyang Gao
Wenfei Guan
Zedong Wang
Zicheng Liu
Lirong Wu
Stan Z. Li

Spatio-temporal predictive learning is a learning paradigm that enables models to learn spatial and temporal patterns by predicting future frames from given past frames in an unsupervised manner. Despite remarkable progress in recent years, a lack of systematic understanding persists due to the diverse settings, complex implementation, and difficult reproducibility. Without standardization, comparisons can be unfair and insights inconclusive. To address this dilemma, we propose OpenSTL, a comprehensive benchmark for spatio-temporal predictive learning that categorizes prevalent approaches into recurrent-based and recurrent-free models. OpenSTL provides a modular and extensible framework implementing various state-of-the-art methods. We conduct standard evaluations on datasets across various domains, including synthetic moving object trajectory, human motion, driving scenes, traffic flow, and weather forecasting. Based on our observations, we provide a detailed analysis of how model architecture and dataset properties affect spatio-temporal predictive learning performance. Surprisingly, we find that recurrent-free models achieve a good balance between efficiency and performance than recurrent models. Thus, we further extend the common MetaFormers to boost recurrent-free spatial-temporal predictive learning. We open-source the code and models at https: //github. com/chengtan9907/OpenSTL.

PDF Details

YNIMG Journal 2021 Journal Article

Structural and functional connectivity mapping of the human corpus callosum organization with white-matter functional networks

Pan Wang
Jianlin Wang
Qin Tang
Tara L. Alvarez
Zedong Wang
Yi-Chia Kung
Ching-Po Lin
Huafu Chen

The corpus callosum serves as a crucial organization for understanding the information integration between the two hemispheres. Our previous study explored the functional connectivity between the corpus callosum and white-matter functional networks (WM-FNs), but the corresponding physical connectivity remains unknown. The current study uses the resting-state fMRI of Human Connectome Project data to identify ten WM-FNs in 108 healthy subjects, and then independently maps the structural and functional connectivity between the corpus callosum and above WM-FNs using the diffusion tensor images (DTI) tractography and resting-state functional connectivity (RSFC). Our results demonstrated that the structural and functional connectivity between the human corpus callosum and WM-FNs have the following high overall correspondence: orbitofrontal WM-FN, DTI map = 89% and RSFC map = 92%; sensorimotor middle WM-FN, DTI map = 47% and RSFC map = 77%; deep WM-FN, DTI map = 50% and RSFC map = 79%; posterior corona radiata WM-FN, DTI map = 82% and RSFC map = 73%. These findings reinforce the notion that the corpus callosum has unique spatial distribution patterns connecting to distinct WM-FNs. However, important differences between the structural and functional connectivity mapping results were also observed, which demonstrated a synergy between DTI tractography and RSFC toward better understanding the information integration of primary and higher-order functional systems in the human brain.

Details DOI