Exploiting Completeness and Uncertainty of Pseudo Labels for Weakly Supervised Video Anomaly Detection
Institute of Information Engineering, CAS(中国科学院信息工程研究所);University of Chinese Academy of Sciences(中国科学院大学);Australian Institute for Machine Learning, The University of Adelaide(阿德莱德大学澳大利亚机器学习研究所);Institute of Computing Technology, CAS(中国科学院计算技术研究所);Peng Cheng Laboratory(鹏城实验室);University of California, Merced(加州大学默塞德分校)
Weakly supervised video anomaly detection aims to identify abnormal events in videos using only video-level labels. Recently, two-stage self-training methods have achieved significant improvements by self-generating pseudo labels and self-refining anomaly scores with these labels. As the pseudo labels play a crucial role, we propose an enhancement framework by exploiting completeness and uncertainty properties for effective self-training. Specifically, we first design a multi-head classification module (each head serves as a classifier) with a diversity loss to maximize the distribution differences of predicted pseudo labels across heads. This encourages the generated pseudo labels to cover as many abnormal events as possible. We then devise an iterative uncertainty pseudo label refinement strategy, which improves not only the initial pseudo labels but also the updated ones obtained by the desired classifier in the second stage. Extensive experimental results demonstrate the proposed method performs favorably against state-of-the-art approaches on the UCF-Crime, TAD, and XD-Violence benchmark datasets.
Learning Event Completeness for Weakly Supervised Video Anomaly Detection
School of Computer Science;Technology, Tongji University, Shanghai, China.;Department of R\&D Data, Microsoft Asia-Pacific Technology CO Ltd, Shanghai, China.
Dynamic Erasing Network Based on Multi-Scale Temporal Features for Weakly Supervised Video Anomaly Detection
Institute of Information Engineering, CAS(中国科学院信息工程研究所);University of Chinese Academy of Sciences(中国科学院大学);Australian Institute for Machine Learning(澳大利亚机器学习研究所);The University of Adelaide(阿德莱德大学);University of California, Merced(加州大学默塞德分校);Institute of Computing Technology, CAS(中国科学院计算技术研究所)
StructCore: Structure-Aware Image-Level Scoring for Training-Free Unsupervised Anomaly Detection
Tsinghua University Shenzhen International Graduate School(清华大学深圳国际研究生院);Ratel Soft;Affiliated Fifth Hospital, Wenzhou Medical University(温州医科大学附属第五医院);Chinese Medicine Guangdong Laboratory(广东中医药实验室)
Shanghai Jiao Tong University(上海交通大学);Shanghai AI Laboratory(上海人工智能实验室);Ruijin Hospital, Shanghai Jiao Tong University School of Medicine(上海交通大学医学院附属瑞金医院)
LGN-Net: Local-Global Normality Network for Video Anomaly Detection
Fudan University(复旦大学);Henan University of Science and Technology(河南科技大学);University of Shanghai for Science and Technology(上海理工大学);Shanghai University of Electric Power(上海电力大学)
GeoMAD: Geometry-Aware Multi-View Anomaly Detection via Deformable Fusion and Distributional Alignment
National Taiwan University(台湾大学);National Taiwan Normal University(台湾师范大学);Mitsubishi Electric Research Laboratories (MERL)(三菱电机研究院);Microsoft Taiwan Corporation(微软台湾公司);VinUniversity;National Taiwan University of Science and Technology(台湾科技大学)
Exploiting Completeness Perception with Diffusion Transformer for Unified 3D MRI Synthesis
School of Engineering, University of Birmingham, UK(伯明翰大学工程学院);William Harvey Research Institute, Queen Mary University London, UK(女王玛丽大学伦敦威廉·哈里维研究所);Barts Heart Centre, St Bartholomew’s Hospital, Barts Health NHS Trust, UK(巴特勒心脏中心,圣巴塞洛缪医院,巴特勒健康 NHS信托);Division of Cardiology, Johns Hopkins University School of Medicine, US(约翰霍普金斯大学医学院心脏病科)
MoniTor: Exploiting Large Language Models with Instruction for Online Video Anomaly Detection
College of Artificial Intelligence, Nanjing University of Aeronautics and Astronautics(人工智能学院,南京航空航天大学);Key Laboratory of Brain-Machine Intelligence Technology, Ministry of Education, China(脑机智能技术重点实验室,教育部,中国)
Exploiting Epistemic Uncertainty of Anatomy Segmentation for Anomaly Detection in Retinal OCT
Medical University Vienna(维也纳医科大学);University of Vienna(维也纳大学);NVIDIA Corporation(英伟达公司)
通过检测医学图像中的相关生物标志物来辅助诊断和治疗指导。虽然监督式深度学习可以对病理区域进行准确分割,但它受到限制,需要对这些区域进行先验定义,大规模注释以及训练集中的代表性患者队列。相反,异常检测不限于病理的特定定义,并且允许在没有注释的情况下对健康样本进行训练。然后,异常区域可以作为生物标志物发现的候选者。关于正常解剖结构的知识带来了用于检测异常的隐含信息。我们建议使用baidu深度学习来利用这一特性,基于这样的假设,即认知的不确定性将与正常训练集的解剖偏差相关。贝叶斯U形网是在一个定义良好的健康环境中训练的,使用现有方法产生的健康解剖结构的弱标签。在测试时,我们使用Monte Carlo dropout捕获模型的认知不确定性估计。一种新的后处理技术,然后应用到利用这些估计,并将其分层的外观平滑斑点形状的分割的异常。我们实验验证了这种方法在视网膜光学相干断层扫描(OCT)图像,使用弱标签的视网膜层。我们的方法在年龄相关性黄斑变性(AMD)病例的独立异常测试集中获得了0.789的Dice指数。所得到的分割允许非常高的准确性,以分离健康和患病的情况下,晚期湿性AMD,干地图状萎缩(GA),糖尿病性黄斑水肿(DME)和视网膜静脉阻塞(RVO)。最后,我们定性地观察到,我们的方法也可以检测正常扫描中的其他偏差,如切割边缘伪影。
Exploiting Autoencoder's Weakness to Generate Pseudo Anomalies
University of Science and Technology(科学技术大学);Electronics and Telecommunications Research Institute(电子通信研究院);University of Luxembourg(卢森堡大学);Mohamed Bin Zayed University of Artificial Intelligence(穆罕默德·本·扎耶德人工智能大学)
Reservoir-enhanced Segment Anything Model for Subsurface Diagnosis
School of Computer Science and Technology, University of Science and Technology of China(中国科学技术大学计算机科学与技术学院);Research Institute of Mine Artificial Intelligence, Chinese Institute of Coal Science(中国煤炭科学研究院矿山人工智能研究院)