Yu Tian is a postdoctoral research fellow at Harvard University. He received his Ph.D. in computer science at the Australian Institute for Machine Learning (AIML), University of Adelaide. Previously, Dr. Tian obtained his bachelor’s degree in computer science with First Class Honours at the University of Adelaide.
ResearchMy main research interests are in the fields of computer vision and medical image analysis, in particular abnormality and rarity learning tasks, such as image/video anomaly detection for surveillance and industrial applications or early detection of diseases.
|One paper is accepted to IEEE Transactions on Medical Imaging (TMI).
|Two papers are accepted to CVPR2024.
|One paper on OoD detection for NLP systems is accepted in COLING2024.
|Two papers are accepted in ICLR2024. One about fairness learning and another about zero-shot anomaly detection.
|One paper about medical anomaly detection is accepted in MICCAI-MLMI.
|4 papers are accepted in ICCV 2023.
Selected Publications* denotes equal contribution; ^ denotes corresponding authorship.
Harvard Glaucoma Fairness: A Retinal Nerve Disease Dataset for Fairness Learning and Fair Identity NormalizationIEEE Transactions on Medical Imaging (
FairSeg: A Large-Scale Medical Image Segmentation Dataset for Fairness Learning Using Segment Anything Model with Fair Error-Bound ScalingInternational Conference on Learning Representations (
AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly DetectionInternational Conference on Learning Representations (
Semantic Role Labeling Guided Out-of-distribution DetectionInternational Conference on Computational Linguistics (
Harvard Glaucoma Detection and Progression: A Multimodal Multitask Dataset and Generalization-Reinforced Semi-Supervised LearningIn Proceedings of the IEEE/CVF international conference on computer vision (
Self-supervised Pseudo Multi-class Pre-training for Unsupervised Anomaly Detection and Segmentation in Medical ImagesMedical Image Analysis (
Pixel-wise Energy-biased Abstention Learning for Anomaly Segmentation on Complex Urban Driving ScenesEuropean Conference on Computer Vision (
ECCV Oral), 2022
Contrastive Transformer-based Multiple Instance Learning for Weakly Supervised Polyp Frame DetectionInternational Conference on Medical Image Computing and Computer-Assisted Intervention (
MICCAI Early Accept), 2022
ACPL: Anti-curriculum Pseudo-labelling for Semi-supervised Medical Image ClassificationIEEE Conference on Computer Vision and Pattern Recognition (
Deep One-Class Classification via Interpolated Gaussian DescriptorIn Thirty-Sixth AAAI Conference on Artificial Intelligence (
AAAI Oral), 2022
Weakly-supervised Video Anomaly Detection with Robust Temporal Feature Magnitude LearningIn Proceedings of the IEEE/CVF international conference on computer vision (
Constrained Contrastive Distribution Learning for Unsupervised Anomaly Detection and Localisation in Medical ImagesIn International Conference on Medical Image Computing and Computer-Assisted Intervention (
Few-shot anomaly detection for polyp frames from colonoscopyIn International Conference on Medical Image Computing and Computer-Assisted Intervention (