Deep neural networks have demonstrated their ability to automatically extract meaningful features from data. However, in supervised learning, information specific to the dataset used for training, but irrelevant to the task at hand, may remain encoded in the extracted representations. This remaining information introduces a domain-specific bias, weakening the generalization performance. In this work, we propose splitting the information into a task-related representation and its complementary context representation. We propose an original method, combining adversarial feature predictors and cyclic reconstruction, to disentangle these two representations in the single-domain supervised case. We then adapt this method to the unsupervised domain adaptation problem, consisting of training a model capable of performing on both a source and a target domain. In particular, our method promotes disentanglement in the target domain, despite the absence of training labels. This enables the isolation of task-specific information from both domains and a projection into a common representation. The task-specific representation allows efficient transfer of knowledge acquired from the source domain to the target domain. In the single-domain case, we demonstrate the quality of our representations on information retrieval tasks and the generalization benefits induced by sharpened task-specific representations. We then validate the proposed method on several classical domain adaptation benchmarks and illustrate the benefits of disentanglement for domain adaptation.
Variantional autoencoder with decremental information bottleneck for disentanglement
Surrey Institute for People Centred AI(萨里以人为本人工智能研究所);University of Surrey(萨里大学);Carnegie Mellon University(卡内基梅隆大学);University of Jinan(济南大学);Mintyea Tech, Inc.(茗茶科技公司)
Structural Disentanglement in Bilinear MLPs via Architectural Inductive Bias
Department of Electrical Engineering, Indian Institute of Technology Roorkee (IIT Roorkee), India;Department of Material Science \& Engineering, IIT Roorkee, India;Mehta Family School of Data Science
Mutual Information Collapse Explains Disentanglement Failure in $β$-VAEs
Department of Mathematics, Miami University, Oxford, OH, USA;Department of Mathematics, Louisiana State University, Baton Rouge, LA, USA;Division of Electrical \& Computer Engineering, Louisiana State University, Baton Rouge, LA, USA
Disentanglement with Factor Quantized Variational Autoencoders
Department of Mathematics and Computer Science, University of Southern Denmark(数学与计算机科学系,丹麦南部大学);Department of Computer Engineering, Istanbul Technical University(计算机工程系,伊斯坦布尔技术大学);Department of AI and Data Engineering, Istanbul Technical University(人工智能与数据工程系,伊斯坦布尔技术大学)
Conceptualizing Embeddings: Sparse Disentanglement for Vision-Language Models
Faculty of Mathematics and Computer Science, Jagiellonian University(雅盖隆大学数学与计算机科学学院);Doctoral School of Exact and Natural Sciences, Jagiellonian University(雅盖隆大学精确与自然科学博士学校);Centre for Credible AI, Warsaw University of Technology(华沙技术大学可信人工智能中心)
Variational Inference of Disentangled Latent Concepts from Unlabeled Observations
IBM Research AI(IBM研究人工智能部门);Yorktown Heights, NY(纽约州约克镇高地(这里注意:根据提取规则,地址信息不应提取,但原HTML中“Affiliation: Yorktown Heights, NY”可能是地址而非机构,结合上下文,正确机构应为“IBM Research AI”,另一个“Affiliation”可能是地址或重复,需确认。但根据提取规则,只提取明确的机构。原文本中明确的机构是“IBM Research AI”,“Yorktown Heights, NY”是地点,非机构。因此正确应为[{)
Structure-aware Unsupervised Tagged-to-Cine MRI Synthesis with Self Disentanglement
Massachusetts General Hospital(麻省总医院);Harvard Medical School(哈佛医学院);Johns Hopkins University(约翰斯·霍普金斯大学);University of Maryland School of Dentistry(马里兰大学牙医学院)
High-Resolution Pelvic MRI Reconstruction Using a Generative Adversarial Network with Attention and Cyclic Loss
Yantai University(烟台大学);Fudan University(复旦大学);Royal Brompton Hospital(皇家布朗普顿医院);Imperial College London(帝国理工学院)
磁共振成像(MRI)是一种重要的医学成像方式,但由于生理条件的限制,其采集速度相当慢。最近,超分辨率方法在加速MRI方面表现出优异的性能。在某些情况下,即使延长扫描时间也难以获得高分辨率图像。因此,我们提出了一种新的超分辨率方法,该方法使用具有循环丢失和注意力机制的生成对抗网络(GAN),以2倍的因子从低分辨率MR图像生成高分辨率MR图像。我们在健康受试者的骨盆图像上实现了我们的模型作为训练和验证数据,而来自患者的数据则用于测试。使用不同的成像序列(包括T2、T2 W SPAIR和mDIXON-W)获得MR数据集。四种方法,即,使用BICUBIC、SRCNN、SRGAN和EDSR进行比较。以结构相似度、峰值信噪比、均方根误差和方差膨胀因子为计算指标,对所提方法的性能进行了评价。各种实验结果表明,我们的方法可以更好地恢复高分辨率MR图像的细节相比,其他方法。此外,重建的高分辨率MR图像可以提供更好的肿瘤患者的病变纹理,这是有希望用于临床诊断。
Semi-supervised Pathology Segmentation with Disentangled Representations
University of Edinburgh(爱丁堡大学);University of California(加利福尼亚大学);Cedars-Sinai Medical Center(西达赛奈医疗中心);University of Essex(埃塞克斯大学);The Alan Turing Institute(艾伦·图灵研究所)