no code implementations • 18 Apr 2024 • Jingyao Wang, Yunhan Tian, Yuxuan Yang, Xiaoxin Chen, Changwen Zheng, Wenwen Qiang
Micro-expressions (MEs) are involuntary movements revealing people's hidden feelings, which has attracted numerous interests for its objectivity in emotion detection.
1 code implementation • 16 Apr 2024 • Jianqi Zhang, Jingyao Wang, Wenwen Qiang, Fanjiang Xu, Changwen Zheng, Fuchun Sun, Hui Xiong
Motivated by these findings, we introduce two new PEs: Temporal Position Encoding (T-PE) for temporal tokens and Variable Positional Encoding (V-PE) for variable tokens.
no code implementations • 3 Mar 2024 • Huijie Guo, Ying Ba, Jie Hu, Lingyu Si, Wenwen Qiang, Lei Shi
Specifically, we update our proposed model through a bi-level optimization mechanism, enabling it to capture comprehensive features.
1 code implementation • 25 Jan 2024 • Jiangmeng Li, Fei Song, Yifan Jin, Wenwen Qiang, Changwen Zheng, Fuchun Sun, Hui Xiong
From the perspective of distribution analyses, we disclose that the intrinsic issues behind the phenomenon are the over-multitudinous conceptual knowledge contained in PLMs and the abridged knowledge for target downstream domains, which jointly result in that PLMs mis-locate the knowledge distributions corresponding to the target domains in the universal knowledge embedding space.
1 code implementation • 21 Dec 2023 • Jiangmeng Li, Yifan Jin, Hang Gao, Wenwen Qiang, Changwen Zheng, Fuchun Sun
To this end, we propose a novel hierarchical topology isomorphism expertise embedded graph contrastive learning, which introduces knowledge distillations to empower GCL models to learn the hierarchical topology isomorphism expertise, including the graph-tier and subgraph-tier.
1 code implementation • 10 Dec 2023 • Jingyao Wang, Yi Ren, Zeen Song, Jianqi Zhang, Changwen Zheng, Wenwen Qiang
However, our experiments reveal an unexpected result: there is negative knowledge transfer between tasks, affecting generalization performance.
no code implementations • 30 Aug 2023 • Hongwei Dong, Fangzhou Han, Lingyu Si, Wenwen Qiang, Lamei Zhang
Based on the constructed SCM, we propose a causal intervention based regularization method to eliminate the negative impact of background on feature semantic learning and achieve background debiased SAR-ATR.
no code implementations • 28 Aug 2023 • Jingyao Wang, Zeen Song, Wenwen Qiang, Changwen Zheng
The long-term goal of machine learning is to learn general visual representations from a small amount of data without supervision, mimicking three advantages of human cognition: i) no need for labels, ii) robustness to data scarcity, and iii) learning from experience.
no code implementations • 21 Aug 2023 • Jiangmeng Li, Hang Gao, Wenwen Qiang, Changwen Zheng
To this end, we rethink the existing multi-view learning paradigm from the perspective of information theory and then propose a novel information theoretical framework for generalized multi-view learning.
1 code implementation • 7 Aug 2023 • Yujie Zhou, Wenwen Qiang, Anyi Rao, Ning Lin, Bing Su, Jiaqi Wang
Specifically, 1) we maximize the MI between visual and semantic space for distribution alignment; 2) we leverage the temporal information for estimating the MI by encouraging MI to increase as more frames are observed.
1 code implementation • 2 Aug 2023 • Jiexin Wang, Yujie Zhou, Wenwen Qiang, Ying Ba, Bing Su, Ji-Rong Wen
Human motion prediction (HMP) has emerged as a popular research topic due to its diverse applications, but it remains a challenging task due to the stochastic and aperiodic nature of future poses.
1 code implementation • 18 Jul 2023 • Jingyao Wang, Wenwen Qiang, Xingzhe Su, Changwen Zheng, Fuchun Sun, Hui Xiong
We obtain three conclusions: (i) there is no universal task sampling strategy that can guarantee the optimal performance of meta-learning models; (ii) over-constraining task diversity may incur the risk of under-fitting or over-fitting during training; and (iii) the generalization performance of meta-learning models are affected by task diversity, task entropy, and task difficulty.
no code implementations • 18 Jul 2023 • Zeen Song, Xingzhe Su, Jingyao Wang, Wenwen Qiang, Changwen Zheng, Fuchun Sun
In recent years, self-supervised learning (SSL) has emerged as a promising approach for extracting valuable representations from unlabeled data.
no code implementations • 17 Jul 2023 • Xingzhe Su, Daixi Jia, Fengge Wu, Junsuo Zhao, Changwen Zheng, Wenwen Qiang
In response, we propose a plug-and-play method named Manifold Guidance Sampling, which is also the first unsupervised method to mitigate bias issue in DDPMs.
no code implementations • 28 Jun 2023 • Lingyu Si, Hongwei Dong, Wenwen Qiang, Junzhi Yu, Wenlong Zhai, Changwen Zheng, Fanjiang Xu, Fuchun Sun
To address this issue, in this paper, we discover the correlation between feature discriminability and dimensional structure (DS) by analyzing and observing features extracted from simple and hard tasks.
no code implementations • 31 May 2023 • Xingzhe Su, Changwen Zheng, Wenwen Qiang, Fengge Wu, Junsuo Zhao, Fuchun Sun, Hui Xiong
This study identifies a previously overlooked issue: GANs exhibit a heightened susceptibility to overfitting on remote sensing images. To address this challenge, this paper analyzes the characteristics of remote sensing images and proposes manifold constraint regularization, a novel approach that tackles overfitting of GANs on remote sensing images for the first time.
no code implementations • 9 Mar 2023 • Xingzhe Su, Wenwen Qiang, Jie Hu, Fengge Wu, Changwen Zheng, Fuchun Sun
Based on this SCM, we theoretically prove that the quality of generated images is positively correlated with the amount of feature information.
no code implementations • 20 Jan 2023 • Hang Gao, Jiangmeng Li, Wenwen Qiang, Lingyu Si, Xingzhe Su, Fengge Wu, Changwen Zheng, Fuchun Sun
By further observing the ramifications of introducing expertise logic into graph representation learning, we conclude that leading the GNNs to learn human expertise can improve the model performance.
1 code implementation • 30 Sep 2022 • Dengsheng Chen, Jie Hu, Wenwen Qiang, Xiaoming Wei, Enhua Wu
In this work, we deep dive into the model's behaviors with skip connections which can be formulated as a learnable Markov chain.
2 code implementations • 16 Sep 2022 • Jiangmeng Li, Wenwen Qiang, Changwen Zheng, Bing Su, Farid Razzak, Ji-Rong Wen, Hui Xiong
To this end, we propose a methodology, specifically consistency and complementarity network (CoCoNet), which avails of strict global inter-view consistency and local cross-view complementarity preserving regularization to comprehensively learn representations from multiple views.
2 code implementations • 16 Sep 2022 • Jiangmeng Li, Wenwen Qiang, Yanan Zhang, Wenyi Mo, Changwen Zheng, Bing Su, Hui Xiong
As a successful approach to self-supervised learning, contrastive learning aims to learn invariant information shared among distortions of the input sample.
1 code implementation • 26 Aug 2022 • Jiangmeng Li, Yanan Zhang, Wenwen Qiang, Lingyu Si, Chengbo Jiao, Xiaohui Hu, Changwen Zheng, Fuchun Sun
To understand the reasons behind this phenomenon, we revisit the learning paradigm of knowledge distillation on the few-shot object detection task from the causal theoretic standpoint, and accordingly, develop a Structural Causal Model.
1 code implementation • 18 Aug 2022 • Hang Gao, Jiangmeng Li, Wenwen Qiang, Lingyu Si, Bing Xu, Changwen Zheng, Fuchun Sun
This observation reveals that there exist confounders in graphs, which may interfere with the model learning semantic information, and current graph representation learning methods have not eliminated their influence.
no code implementations • 29 Jun 2022 • Wenwen Qiang, Jiangmeng Li, Changwen Zheng, Bing Su, Hui Xiong
Contrastive learning (CL)-based self-supervised learning models learn visual representations in a pairwise manner.
no code implementations • 23 May 2022 • Jiangmeng Li, Wenyi Mo, Wenwen Qiang, Bing Su, Changwen Zheng
Vision-language models are pre-trained by aligning image-text pairs in a common space so that the models can deal with open-set visual concepts by learning semantic information from textual labels.
2 code implementations • 10 Mar 2022 • Jiangmeng Li, Wenwen Qiang, Changwen Zheng, Bing Su, Hui Xiong
We perform a meta learning technique to build the augmentation generator that updates its network parameters by considering the performance of the encoder.
no code implementations • 8 Mar 2022 • Wenwen Qiang, Jiangmeng Li, Changwen Zheng, Bing Su, Hui Xiong
We conduct theoretical analysis on the robustness of the proposed RLPGA and prove that the robust informative-theoretic-based loss and the local preserving module are beneficial to reduce the empirical risk of the target domain.
1 code implementation • 11 Jan 2022 • Hang Gao, Jiangmeng Li, Wenwen Qiang, Lingyu Si, Fuchun Sun, Changwen Zheng
To this end, we propose a novel approach to learning a graph augmenter that can generate an augmentation with uniformity and informativeness.
no code implementations • 29 Sep 2021 • Wenwen Qiang, Jiangmeng Li, Jie Hu, Bing Su, Changwen Zheng, Hui Xiong
In this paper, we give an analysis of the existing representation learning framework of unsupervised domain adaptation and show that the learned feature representations of the source domain samples are with discriminability, compressibility, and transferability.
no code implementations • 6 Sep 2021 • Jiangmeng Li, Wenwen Qiang, Hang Gao, Bing Su, Farid Razzak, Jie Hu, Changwen Zheng, Hui Xiong
To this end, we rethink the existing multi-view learning paradigm from the information theoretical perspective and then propose a novel information theoretical framework for generalized multi-view learning.