1 code implementation • 5 Mar 2024 • Fei Zhu, Xu-Yao Zhang, Zhen Cheng, Cheng-Lin Liu
Reliable confidence estimation is a challenging yet fundamental requirement in many risk-sensitive applications.
no code implementations • 4 Mar 2024 • Fei Zhu, Shijie Ma, Zhen Cheng, Xu-Yao Zhang, Zhaoxiang Zhang, Cheng-Lin Liu
This paper aims to provide a comprehensive introduction to the emerging open-world machine learning paradigm, to help researchers build more powerful AI systems in their respective fields, and to promote the development of artificial general intelligence.
1 code implementation • 22 Nov 2023 • Zhen Cheng, Xu-Yao Zhang, Cheng-Lin Liu
Previous methods mostly take post-training score transformation or hybrid models to ensure low scores on OOD inputs while separating known classes.
1 code implementation • 18 Jul 2023 • Shijie Ma, Fei Zhu, Zhen Cheng, Xu-Yao Zhang
By distilling both InD samples and outliers, the condensed datasets are capable to train models competent in both InD classification and OOD detection.
1 code implementation • CVPR 2023 • Xiao Zhou, Yujie Zhong, Zhen Cheng, Fan Liang, Lin Ma
To address this problem, we propose a novel loss paradigm termed Sparse Pairwise (SP) loss that only leverages few appropriate pairs for each class in a mini-batch, and empirically demonstrate that it is sufficient for the ReID tasks.
1 code implementation • CVPR 2023 • Fei Zhu, Zhen Cheng, Xu-Yao Zhang, Cheng-Lin Liu
Reliable confidence estimation for deep neural classifiers is a challenging yet fundamental requirement in high-stakes applications.
1 code implementation • 25 Mar 2023 • Jiacheng Li, Chang Chen, Zhen Cheng, Zhiwei Xiong
However, the size of a single LUT grows exponentially with the increase of its indexing capacity, which restricts its receptive field and thus the performance.
1 code implementation • 6 Mar 2023 • Fei Zhu, Zhen Cheng, Xu-Yao Zhang, Cheng-Lin Liu
We investigate this problem and reveal that popular confidence calibration methods often lead to worse confidence separation between correct and incorrect samples, making it more difficult to decide whether to trust a prediction or not.
no code implementations • 2 Mar 2023 • Zhen Cheng, Fei Zhu, Xu-Yao Zhang, Cheng-Lin Liu
Detecting Out-of-distribution (OOD) inputs have been a critical issue for neural networks in the open world.
no code implementations • 6 Nov 2022 • Zhen Cheng, Tao Wang, Yong Li, Fenglong Song, Chang Chen, Zhiwei Xiong
To solve this problem, we propose a learning-based data synthesis approach to learn the properties of real-world SDRTVs by integrating several tone mapping priors into both network structures and loss functions.
no code implementations • CVPR 2022 • Xihao Chen, Zhiwei Xiong, Zhen Cheng, Jiayong Peng, Yueyi Zhang, Zheng-Jun Zha
Interestingly, we find that, although a stereo matching network trained with the photometric loss is not optimal, its feature extractor can produce degradation-agnostic and matching-specific features.
2 code implementations • NeurIPS 2021 • Fei Zhu, Zhen Cheng, Xu-Yao Zhang, Cheng-Lin Liu
Deep learning systems typically suffer from catastrophic forgetting of past knowledge when acquiring new skills continually.
no code implementations • 29 Sep 2021 • Zhen Cheng, Fei Zhu, Xu-Yao Zhang, Cheng-Lin Liu
Comprehensive experiments demonstrate that FSR is effective to alleviate the dominance of larger eigenvalues and improve adversarial robustness on different datasets.
no code implementations • CVPR 2021 • Zeyu Xiao, Xueyang Fu, Jie Huang, Zhen Cheng, Zhiwei Xiong
In this paper, we aim to improve the performance of compact VSR networks without changing their original architectures, through a knowledge distillation approach that transfers knowledge from a complicated VSR network to a compact one.
no code implementations • CVPR 2021 • Zhen Cheng, Zhiwei Xiong, Chang Chen, Dong Liu, Zheng-Jun Zha
To fill this gap, we propose a zero-shot learning framework for light field SR, which learns a mapping to super-resolve the reference view with examples extracted solely from the input low-resolution light field itself.
no code implementations • 9 Nov 2019 • Zhen Cheng, Zaixiang Zheng, Xin-yu Dai, Shu-Jian Huang, Jia-Jun Chen
Intuitively, NLI should rely more on multiple perspectives to form a holistic view to eliminate bias.