1 code implementation • 16 May 2024 • Feiran Li, Qianqian Xu, Shilong Bao, Zhiyong Yang, Runmin Cong, Xiaochun Cao, Qingming Huang
This paper explores the size-invariance of evaluation metrics in Salient Object Detection (SOD), especially when multiple targets of diverse sizes co-exist in the same image.
1 code implementation • 15 May 2024 • Cong Hua, Qianqian Xu, Shilong Bao, Zhiyong Yang, Qingming Huang
This paper explores a novel multi-modal alternating learning paradigm pursuing a reconciliation between the exploitation of uni-modal features and the exploration of cross-modal interactions.
1 code implementation • 13 May 2024 • Zhiyong Yang, Qianqian Xu, Zitai Wang, Sicong Li, Boyu Han, Shilong Bao, Xiaochun Cao, Qingming Huang
Traditional methods predominantly use a Mixture-of-Expert (MoE) approach, targeting a few fixed test label distributions that exhibit substantial global variations.
Ranked #1 on Test Agnostic Long-Tailed Learning on CIFAR-10-LT
1 code implementation • NeurIPS 2023 • Siran Dai, Qianqian Xu, Zhiyong Yang, Xiaochun Cao, Qingming Huang
To tackle this challenge, methodically we propose an instance-wise surrogate loss of Distributionally Robust AUC (DRAUC) and build our optimization framework on top of it.
no code implementations • 12 Oct 2023 • Peifeng Gao, Qianqian Xu, Yibo Yang, Peisong Wen, Huiyang Shao, Zhiyong Yang, Bernard Ghanem, Qingming Huang
While there have been extensive studies on optimization characteristics showing the global optimality of neural collapse, little research has been done on the generalization behaviors during the occurrence of NC.
1 code implementation • NeurIPS 2023 • Zitai Wang, Qianqian Xu, Zhiyong Yang, Yuan He, Xiaochun Cao, Qingming Huang
However, existing generalization analysis of such losses is still coarse-grained and fragmented, failing to explain some empirical results.
1 code implementation • 7 Oct 2023 • Zitai Wang, Qianqian Xu, Zhiyong Yang, Yuan He, Xiaochun Cao, Qingming Huang
However, existing generalization analysis of such losses is still coarse-grained and fragmented, failing to explain some empirical results.
Ranked #6 on Long-tail Learning on CIFAR-10-LT (ρ=10)
1 code implementation • journal 2023 • Junyu Chen, Qianqian Xu, Zhiyong Yang, Ke Ma, Xiaochun Cao, Qingming Huang
For the motif-based node representation learning process, we propose a Motif Coarsening strategy for incorporating motif structure into the graph representation learning process.
1 code implementation • 30 Sep 2023 • Zhiyong Yang, Yuelong Zhu, Xiaoqin Zeng, Jun Zong, Xiuheng Liu, Ran Tao, Xiaofei Cong, YuFeng Yu
First, we built a river ice semantic segmentation dataset IPC_RI_SEG using a fixed camera and covering the entire ice melting process of the river.
1 code implementation • TPAMI 2023 • Zhiyong Yang, Qianqian Xu, Shilong Bao, Peisong Wen, Xiaochun Cao, Qingming Huang
We propose a new result that not only addresses the interdependency issue but also brings a much sharper bound with weaker assumptions about the loss function.
2 code implementations • TPAMI 2023 • Zhiyong Yang, Qianqian Xu, Wenzheng Hou, Shilong Bao, Yuan He, Xiaochun Cao, Qingming Huang
On top of this, we can show that: 1) Under mild conditions, AdAUC can be optimized equivalently with score-based or instance-wise-loss-based perturbations, which is compatible with most of the popular adversarial example generation methods.
no code implementations • 18 Apr 2023 • Peifeng Gao, Qianqian Xu, Peisong Wen, Huiyang Shao, Zhiyong Yang, Qingming Huang
Out of curiosity about the symmetry of Grassmannian Frame, we conduct experiments to explore if models with different Grassmannian Frames have different performance.
no code implementations • ICCV 2023 • Huiyang Shao, Qianqian Xu, Peisong Wen, Peifeng Gao, Zhiyong Yang, Qingming Huang
Finally, experimental results support the effectiveness of the proposed framework in terms of both mural synthesis and restoration.
1 code implementation • 22 Oct 2022 • Zitai Wang, Qianqian Xu, Zhiyong Yang, Yuan He, Xiaochun Cao, Qingming Huang
In this paper, a systematic analysis reveals that most existing metrics are essentially inconsistent with the aforementioned goal of OSR: (1) For metrics extended from close-set classification, such as Open-set F-score, Youden's index, and Normalized Accuracy, a poor open-set prediction can escape from a low performance score with a superior close-set prediction.
2 code implementations • NeurIPS 2022 • Huiyang Shao, Qianqian Xu, Zhiyong Yang, Shilong Bao, Qingming Huang
sample size and a slow convergence rate, especially for TPAUC.
1 code implementation • Conference 2022 • Junyu Chen, Qianqian Xu, Zhiyong Yang, Ke Ma, Xiaochun Cao, Qingming Huang
To attack this problem, we propose a recursive meta-learning model with the user's behavior sequence prediction as a separate training task.
2 code implementations • Proceedings of the 30th ACM International Conference on Multimedia 2022 • Junyu Chen, Qianqian Xu, Zhiyong Yang, Xiaochun Cao, Qingming Huang
We develop a multi-class AUC optimization work to deal with the class imbalance problem.
1 code implementation • NeurIPS 2023 • Shilong Bao, Qianqian Xu, Zhiyong Yang, Yuan He, Xiaochun Cao, Qingming Huang
Collaborative Metric Learning (CML) has recently emerged as a popular method in recommendation systems (RS), closing the gap between metric learning and Collaborative Filtering.
1 code implementation • 27 Sep 2022 • Peisong Wen, Qianqian Xu, Zhiyong Yang, Yuan He, Qingming Huang
Stochastic optimization of the Area Under the Precision-Recall Curve (AUPRC) is a crucial problem for machine learning.
no code implementations • 26 Sep 2022 • Yangbangyan Jiang, Xiaodan Li, Yuefeng Chen, Yuan He, Qianqian Xu, Zhiyong Yang, Xiaochun Cao, Qingming Huang
In recent years, great progress has been made to incorporate unlabeled data to overcome the inefficiently supervised problem via semi-supervised learning (SSL).
1 code implementation • 3 Sep 2022 • Zitai Wang, Qianqian Xu, Zhiyong Yang, Yuan He, Xiaochun Cao, Qingming Huang
Finally, the experimental results on four benchmark datasets validate the effectiveness of our proposed framework.
no code implementations • 28 Jun 2022 • Tianwei Cao, Qianqian Xu, Zhiyong Yang, Qingming Huang
In this paper, we regard user interest modeling as a feature selection problem, which we call user interest selection.
1 code implementation • 24 Jun 2022 • Zongsheng Cao, Qianqian Xu, Zhiyong Yang, Qingming Huang
To address this issue, we propose a new regularizer, namely, Equivariance Regularizer (ER), which can suppress overfitting by leveraging the implicit semantic information.
no code implementations • ICML 2022 • Wenzheng Hou, Qianqian Xu, Zhiyong Yang, Shilong Bao, Yuan He, Qingming Huang
Our analysis differs from the existing studies since the algorithm is asked to generate adversarial examples by calculating the gradient of a min-max problem.
1 code implementation • 24 Jun 2022 • Zongsheng Cao, Qianqian Xu, Zhiyong Yang, Xiaochun Cao, Qingming Huang
Knowledge graph (KG) embeddings have shown great power in learning representations of entities and relations for link prediction tasks.
no code implementations • TPAMI 2022 • Shilong Bao, Qianqian Xu, Zhiyong Yang, Xiaochun Cao, Qingming Huang
However, in this work, by taking a theoretical analysis, we find that negative sampling would lead to a biased estimation of the generalization error.
1 code implementation • TPAMI 2022 • Zhiyong Yang, Qianqian Xu, Shilong Bao, Yuan He, Xiaochun Cao, Qingming Huang
The critical challenge along this course lies in the difficulty of performing gradient-based optimization with end-to-end stochastic training, even with a proper choice of surrogate loss.
no code implementations • NeurIPS 2021 • Peisong Wen, Qianqian Xu, Zhiyong Yang, Yuan He, Qingming Huang
To leverage high performance under low FPRs, we consider an alternative metric for multipartite ranking evaluating the True Positive Rate (TPR) at a given FPR, denoted as TPR@FPR.
1 code implementation • ACM MM 2021 2021 • Zitai Wang, Qianqian Xu, Zhiyong Yang, Xiaochun Cao, Qingming Huang
As the core of the framework, the iterative relabeling module exploits the self-training principle to dynamically generate pseudo labels for user preferences.
1 code implementation • ACM MM 2021 2021 • Qianxiu Hao, Qianqian Xu, Zhiyong Yang, Qingming Huang
Heterogeneous information networks (HINs) have become a popular tool to capture complicated user-item relationships in recommendation problems in recent years.
2 code implementations • MM '21: Proceedings of the 29th ACM International Conference on Multimedia 2021 • Qianxiu Hao, Qianqian Xu, Zhiyong Yang, Qingming Huang
To balance overall recommendation performance and fairness, prevalent solutions apply fairness constraints or regularizations to enforce equality of certain performance across different subgroups.
no code implementations • TPAMI 2021 • Zhiyong Yang, Qianqian Xu, Shilong Bao, Xiaochun Cao, Qingming Huang
Our foundation is based on the M metric, which is a well-known multiclass extension of AUC.
1 code implementation • ICML 2021 • Zhiyong Yang, Qianqian Xu, Shilong Bao, Yuan He, Xiaochun Cao, Qingming Huang
The critical challenge along this course lies in the difficulty of performing gradient-based optimization with end-to-end stochastic training, even with a proper choice of surrogate loss.
no code implementations • NeurIPS 2021 • Peisong Wen, Qianqian Xu, Zhiyong Yang, Yuan He, Qingming Huang
To leverage high performance under low FPRs, we consider an alternative metric for multipartite ranking evaluating the True Positive Rate (TPR) at a given FPR, denoted as TPR@FPR.
1 code implementation • CVPR 2021 • Peisong Wen, Qianqian Xu, Yangbangyan Jiang, Zhiyong Yang, Yuan He, Qingming Huang
Targeting at (a), we propose a two-level modality alignment loss where both global and local information are considered.
no code implementations • 29 Apr 2020 • Zhiyong Yang, Qianqian Xu, Xiaochun Cao, Qingming Huang
To this end, we propose a novel multi-task learning method called Task-Feature Collaborative Learning (TFCL).
1 code implementation • NeurIPS 2019 • Yangbangyan Jiang, Qianqian Xu, Zhiyong Yang, Xiaochun Cao, Qingming Huang
Instead of transforming all the samples into a joint modality-independent space, our framework learns the mappings across individual modal spaces by virtue of cycle-consistency.
1 code implementation • NeurIPS 2019 • Zhiyong Yang, Qianqian Xu, Yangbangyan Jiang, Xiaochun Cao, Qingming Huang
Different from most of the previous work, pursuing the Block-Diagonal structure of LTAM (assigning latent tasks to output tasks) alleviates negative transfer via collaboratively grouping latent tasks and output tasks such that inter-group knowledge transfer and sharing is suppressed.
1 code implementation • ACM MM 2019 • Shilong Bao, Qianqian Xu, Ke Ma, Zhiyong Yang, Xiaochun Cao, Qingming Huang
From the margin theory point-of-view, we then propose a generalization enhancement scheme for sparse and insufficient labels via optimizing the margin distribution.
1 code implementation • NeurIPS 2019 • Qianqian Xu, Xinwei Sun, Zhiyong Yang, Xiaochun Cao, Qingming Huang, Yuan YAO
In this paper, instead of learning a global ranking which is agreed with the consensus, we pursue the tie-aware partial ranking from an individualized perspective.
no code implementations • 18 Jun 2019 • Zhiyong Yang, Qianqian Xu, Xiaochun Cao, Qingming Huang
Traditionally, most of the existing attribute learning methods are trained based on the consensus of annotations aggregated from a limited number of annotators.
no code implementations • CVPR 2019 • Qianqian Xu, Zhiyong Yang, Yangbangyan Jiang, Xiaochun Cao, Qingming Huang, Yuan YAO
The problem of estimating subjective visual properties (SVP) of images (e. g., Shoes A is more comfortable than B) is gaining rising attention.
1 code implementation • 5 Dec 2018 • Ke Ma, Qianqian Xu, Zhiyong Yang, Xiaochun Cao
To address the issue of insufficient training samples, we propose a margin distribution learning paradigm for ordinal embedding, entitled Distributional Margin based Ordinal Embedding (\textit{DMOE}).
no code implementations • 29 Jul 2018 • Qianqian Xu, Jiechao Xiong, Xinwei Sun, Zhiyong Yang, Xiaochun Cao, Qingming Huang, Yuan YAO
A preference order or ranking aggregated from pairwise comparison data is commonly understood as a strict total order.
no code implementations • 18 Nov 2017 • Zhiyong Yang, Qianqian Xu, Xiaochun Cao, Qingming Huang
However, both categories ignore the joint effect of the two mentioned factors: the personal diversity with respect to the global consensus; and the intrinsic correlation among multiple attributes.