Author name cluster

Sheng Guo

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.

9 papers

2 author rows

AAAI Conference 2026 Conference Paper

Trimming the Fat: Redundancy-Aware Acceleration Framework for DGNNs

Renhong Huang
Yuxuan Cao
Yi Li
Junwei Hu
Zihua Xiong
Shuai Fang
Sheng Guo
Bo Zheng

Temporal graphs are essential for modeling complex real-world systems, such as social interactions, financial transactions, and recommendation systems, but the high computational cost and model complexity of dynamic graph neural networks (DGNNs) pose significant challenges for practical deployment. Although various pruning and sampling techniques have proven effective in accelerating static GNNs, they fall short in dynamic settings due to temporal dependencies in evolving graph structures. To address these challenges, we propose TrimDG, a general framework that accelerates DGNNs by eliminating both static and runtime redundancies. For static redundancy, we introduce a novel node influence metric, Temporal Personalized PageRank (TPP), to prune less informative nodes, and employ temporal binning to remove redundant events. For runtime redundancy during training, we develop an adaptive sampling strategy guided by graph information bottleneck and further reduce sampling frequency through temporal batch selector and sampling cache. Theoretical analysis supports our design, and experiments on real-world datasets show that TrimDG reduces runtime by an average of 83.49% across diverse DGNN backbones, while maintaining strong predictive performance, demonstrating both its efficiency and generalizability.

PDF Details DOI

YNIMG Journal 2025 Journal Article

Atypical hierarchical brain connectivity in autism: Insights from stepwise causal analysis using Liang information flow

Shan Sun
Fei Wang
Fen Xu
Yufeng Deng
Jiwang Ma
Kai Chen
Sheng Guo
X. San Liang

Autism spectrum disorder (ASD) is associated with atypical brain connectivity, yet its hierarchical organization remains underexplored. In this study, we applied the Liang information flow method to analyze stepwise causal functional connectivity in ASD, offering a novel approach to understanding how different brain networks interact. Using resting-state fMRI data from ASD individuals and healthy controls, we observed significant alterations in both positive and negative causal connections across the ventral attention network, limbic network, frontal-parietal network, and default mode network. These disruptions were detected at multiple hierarchical levels, indicating changes in communication patterns across brain regions. By leveraging features of hierarchical causal connectivity, we achieved high classification accuracy between ASD and healthy individuals. Additionally, changes in network node degrees were found to correlate with ASD clinical symptoms, particularly social and communication behaviors. Our findings provide new insights into disrupted hierarchical brain connectivity in ASD and demonstrate the potential of this approach for distinguishing ASD from typical development.

Details DOI

IROS Conference 2024 Conference Paper

3D Affordance Keypoint Detection for Robotic Manipulation

Zhiyang Liu
Ruiteng Zhao
Lei Zhou
Chengran Yuan
Yuwei Wu 0002
Sheng Guo
Zhengshen Zhang
Chenchen Liu

This paper presents a novel approach for affordance-informed robotic manipulation by introducing 3D keypoints to enhance the understanding of object parts’ functionality. The proposed approach provides direct information about what the potential use of objects is, as well as guidance on where and how a manipulator should engage, whereas conventional methods treat affordance detection as a semantic segmentation task, focusing solely on answering the what question. To address this gap, we propose a Fusion-based Affordance Keypoint Network (FAKP-Net) by introducing 3D keypoint quadruplet that harnesses the synergistic potential of RGB and Depth image to provide information on execution position, direction, and extent. Benchmark testing demonstrates that FAKP-Net outperforms existing models by significant margins in affordance segmentation task and keypoint detection task. Real-world experiments also showcase the reliability of our method in accomplishing manipulation tasks with previously unseen objects. Our source code and video demo will be public.

Details

ICML Conference 2024 Conference Paper

Mind the Boundary: Coreset Selection via Reconstructing the Decision Boundary

Shuo Yang 0006
Zhe Cao 0001
Sheng Guo
Ruiheng Zhang 0001
Ping Luo 0002
Shengping Zhang
Liqiang Nie

Existing paradigms of pushing the state of the art require exponentially more training data in many fields. Coreset selection seeks to mitigate this growing demand by identifying the most efficient subset of training data. In this paper, we delve into geometry-based coreset methods and preliminarily link the geometry of data distribution with models’ generalization capability in theoretics. Leveraging these theoretical insights, we propose a novel coreset construction method by selecting training samples to reconstruct the decision boundary of a deep neural network learned on the full dataset. Extensive experiments across various popular benchmarks demonstrate the superiority of our method over multiple competitors. For the first time, our method achieves a 50% data pruning rate on the ImageNet-1K dataset while sacrificing less than 1% in accuracy. Additionally, we showcase and analyze the remarkable cross-architecture transferability of the coresets derived from our approach.

Details

AAAI Conference 2023 Conference Paper

CoMAE: Single Model Hybrid Pre-training on Small-Scale RGB-D Datasets

Jiange Yang
Sheng Guo
Gangshan Wu
Limin Wang

Current RGB-D scene recognition approaches often train two standalone backbones for RGB and depth modalities with the same Places or ImageNet pre-training. However, the pre-trained depth network is still biased by RGB-based models which may result in a suboptimal solution. In this paper, we present a single-model self-supervised hybrid pre-training framework for RGB and depth modalities, termed as CoMAE. Our CoMAE presents a curriculum learning strategy to unify the two popular self-supervised representation learning algorithms: contrastive learning and masked image modeling. Specifically, we first build a patch-level alignment task to pre-train a single encoder shared by two modalities via cross-modal contrastive learning. Then, the pre-trained contrastive encoder is passed to a multi-modal masked autoencoder to capture the finer context features from a generative perspective. In addition, our single-model design without requirement of fusion module is very flexible and robust to generalize to unimodal scenario in both training and testing phases. Extensive experiments on SUN RGB-D and NYUDv2 datasets demonstrate the effectiveness of our CoMAE for RGB and depth representation learning. In addition, our experiment results reveal that CoMAE is a data-efficient representation learner. Although we only use the small-scale and unlabeled training set for pre-training, our CoMAE pre-trained models are still competitive to the state-of-the-art methods with extra large-scale and supervised RGB dataset pre-training. Code will be released at https://github.com/MCG-NJU/CoMAE.

PDF Details DOI

TMLR Journal 2022 Journal Article

Completeness and Coherence Learning for Fast Arbitrary Style Transfer

Zhijie Wu
Chunjin Song
Guanxiong Chen
Sheng Guo
Weilin Huang

Style transfer methods put a premium on two objectives: (1) completeness which encourages the encoding of a complete set of style patterns; (2) coherence which discourages the production of spurious artifacts not found in input styles. While existing methods pursue the two objectives either partially or implicitly, we present the Completeness and Coherence Network (CCNet) which jointly learns completeness and coherence components and rejects their incompatibility, both in an explicit manner. Specifically, we develop an attention mechanism integrated with bi-directional softmax operations for explicit imposition of the two objectives and for their collaborative modelling. We also propose CCLoss as a quantitative measure for evaluating the quality of a stylized image in terms of completeness and coherence. Through an empirical evaluation, we demonstrate that compared with existing methods, our method strikes a better tradeoff between computation costs, generalization ability and stylization quality.

PDF Details

AAAI Conference 2022 Conference Paper

InsCLR: Improving Instance Retrieval with Self-Supervision

Zelu Deng
Yujie Zhong
Sheng Guo
Weilin Huang

This work aims at improving instance retrieval with selfsupervision. We find that fine-tuning using the recently developed self-supervised learning (SSL) methods, such as Sim- CLR and MoCo, fails to improve the performance of instance retrieval. In this work, we identify that the learnt representations for instance retrieval should be invariant to large variations in viewpoint and background etc. , whereas self-augmented positives applied by the current SSL methods can not provide strong enough signals for learning robust instance-level representations. To overcome this problem, we propose InsCLR, a new SSL method that builds on the instance-level contrast, to learn the intra-class invariance by dynamically mining meaningful pseudo positive samples from both mini-batches and a memory bank during training. Extensive experiments demonstrate that InsCLR achieves similar or even better performance than the state-ofthe-art SSL methods on instance retrieval. Code is available at https: //github. com/zeludeng/insclr.

PDF Details

AAAI Conference 2021 Conference Paper

Unchain the Search Space with Hierarchical Differentiable Architecture Search

Guanting Liu
Yujie Zhong
Sheng Guo
Matthew R. Scott
Weilin Huang

Differentiable architecture search (DAS) has made great progress in searching for high-performance architectures with reduced computational cost. However, DAS-based methods mainly focus on searching for a repeatable cell structure, which is then stacked sequentially in multiple stages to form the networks. This configuration significantly reduces the search space, and ignores the importance of connections between the cells. To overcome this limitation, in this paper, we propose a Hierarchical Differentiable Architecture Search (H-DAS) that performs architecture search both at the cell level and at the stage level. Specifically, the cell-level search space is relaxed so that the networks can learn stage-specific cell structures. For the stage-level search, we systematically study the architectures of stages, including the number of cells in each stage and the connections between the cells. Based on insightful observations, we design several search rules and losses, and mange to search for better stage-level architectures. Such hierarchical search space greatly improves the performance of the networks without introducing expensive search cost. Extensive experiments on CIFAR10 and ImageNet demonstrate the effectiveness of the proposed H- DAS. Moreover, the searched stage-level architectures can be combined with the cell structures searched by existing DAS methods to further boost the performance. Code is available at: https: //github. com/msight-tech/research-HDAS

PDF Details

AAAI Conference 2020 Conference Paper

Knowledge Integration Networks for Action Recognition

Shiwen Zhang
Sheng Guo
Limin Wang
Weilin Huang
Matthew Scott

In this work, we propose Knowledge Integration Networks (referred as KINet) for video action recognition. KINet is capable of aggregating meaningful context features which are of great importance to identifying an action, such as human information and scene context. We design a three-branch architecture consisting of a main branch for action recognition, and two auxiliary branches for human parsing and scene recognition which allow the model to encode the knowledge of human and scene for action recognition. We explore two pre-trained models as teacher networks to distill the knowledge of human and scene for training the auxiliary tasks of KINet. Furthermore, we propose a two-level knowledge encoding mechanism which contains a Cross Branch Integration (CBI) module for encoding the auxiliary knowledge into medium-level convolutional features, and an Action Knowledge Graph (AKG) for effectively fusing high-level context information. This results in an end-to-end trainable framework where the three tasks can be trained collaboratively, allowing the model to compute strong context knowledge ef- ﬁciently. The proposed KINet achieves the state-of-the-art performance on a large-scale action recognition benchmark Kinetics-400, with a top-1 accuracy of 77. 8%. We further demonstrate that our KINet has strong capability by transferring the Kinetics-trained model to UCF-101, where it obtains 97. 8% top-1 accuracy.

PDF Details