Timezone: »

Detecting Out-of-distribution Data through In-distribution Class Prior
Xue JIANG · Feng Liu · zhen fang · Hong Chen · Tongliang Liu · Feng Zheng · Bo Han

Tue Jul 25 02:00 PM -- 04:30 PM (PDT) @ Exhibit Hall 1 #633

Given a pre-trained in-distribution (ID) model, the inference-time out-of-distribution (OOD) detection aims to recognize OOD data during the inference stage. However, some representative methods share an unproven assumption that the probability that OOD data belong to every ID class should be the same, i.e., these OOD-to-ID probabilities actually form a uniform distribution. In this paper, we show that this assumption makes the above methods incapable when the ID model is trained with class-imbalanced data.Fortunately, by analyzing the causal relations between ID/OOD classes and features, we identify several common scenarios where the OOD-to-ID probabilities should be the ID-class-prior distribution and propose two strategies to modify existing inference-time detection methods: 1) replace the uniform distribution with the ID-class-prior distribution if they explicitly use the uniform distribution; 2) otherwise, reweight their scores according to the similarity between the ID-class-prior distribution and the softmax outputs of the pre-trained model. Extensive experiments show that both strategies can improve the OOD detection performance when the ID model is pre-trained with imbalanced data, reflecting the importance of ID-class prior in OOD detection.

Author Information

Xue JIANG (Sustech&HKBU)
Feng Liu (University of Melbourne/RIKEN-AIP)

I am a machine learning researcher with research interests in hypothesis testing and trustworthy machine learning. I am currently an Assistant Professor in Statistics (Data Science) at the School of Mathematics and Statistics, The University of Melbourne, Australia. We are also running the Trustworthy Machine Learning and Reasoning (TMLR) Lab where I am one of co-directors (see this page for details). In addition, I am a Visiting Scientist at RIKEN-AIP, Japan, and a Visting Fellow at DeSI Lab, Australian Artificial Intelligence Institute, University of Technology Sydney. I was the recipient of the Australian Laureate postdoctoral fellowship. I received my Ph.D. degree in computer science at the University of Technology Sydney in 2020, advised by Dist. Prof. Jie Lu and Prof. Guangquan Zhang. I was a research intern at the RIKEN-AIP, working on the robust domain adaptation project with Prof. Masashi Sugiyama, Dr. Gang Niu and Dr. Bo Han. I visited Gatsby Computational Neuroscience Unit at UCL and worked on the hypothesis testing project with Prof. Arthur Gretton, Dr. Danica J. Sutherland and Dr. Wenkai Xu. I have received the Outstanding Paper Award of NeurIPS (2022), the Outstanding Reviewer Award of NeurIPS (2021), the Outstanding Reviewer Award of ICLR (2021), the UTS-FEIT HDR Research Excellence Award (2019). My publications are mainly distributed in high-quality journals or conferences, such as Nature Communications, IEEE-TPAMI, IEEE-TNNLS, IEEE-TFS, NeurIPS, ICML, ICLR, KDD, IJCAI, and AAAI. I have served as a senior program committee (SPC) member for IJCAI, ECAI and program committee (PC) members for NeurIPS, ICML, ICLR, AISTATS, ACML, AAAI and so on. I also serve as reviewers for many academic journals, such as JMLR, IEEE-TPAMI, IEEE-TNNLS, IEEE-TFS and so on.

zhen fang (AAII UTS)
Hong Chen (Huazhong Agricultural University)
Tongliang Liu (The University of Sydney)
Feng Zheng (Southern University of Science and Technology)

Feng Zheng is currently an Associate Researcher at Southern University of Science and Technology (SUSTech) in China. His research interests include machine learning (ML), computer vision (CV) and human-computer interaction (HCI). He received a Ph.D. from the University of Sheffield, UK. Before joining SUSTech, he worked as a senior researcher at Tencent YouTu Lab in Shanghai, China. Prior to this, he worked as a postdoctoral researcher at the University of Pittsburgh, USA and as an assistant research professor at the Shenzhen Institute of Advanced Technology, CAS. In terms of academic research, he has published 101 academic papers in international top journals and conferences, including IEEE TPAMI/TITS/TIP, AAAI, NeuIPS, CVPR, and ECCV. Two papers are listed as the highly cited papers according to Web of Science. In addition, he currently severs as an associate editor for IET Image Processing, a local co-chair for IEEE ICME, an area chair for ACM MM, a member of the program committee for several top international AI conferences, including ICLR, ICML, AAAI, IJCAI, NeuIPS, KDD, UAI, etc., as well as reviewers for several leading journals including IEEE TNNLS/TCSVT/TMM, IS, etc. In terms of system developments, he has successfully designed several HCI systems and been granted 5 patents for HCI related technologies. The systems have been reported by several mainstream media (CCTV) and transferred into companies such as Huawei and Skyworth.


More from the Same Authors