Interaction representation-based subspace learning for domain adaptation

Chen, Jia jia; Li, Ao

doi:10.4108/eai.27-8-2020.2296559

Cited by 11 publications

(13 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Despite some trials of hyperparameters tuning, as shown in Table 1, their results were still comparable or worse than no fine-tuning. This may be due to the fact that most DA approaches were designed for classification tasks, which may not be directly applicable to our regression task [3]. This further verifies the value of our work.…”

Section: Volume-sampled Imagesmentioning

confidence: 54%

Adaptive 3D Localization of 2D Freehand Ultrasound Brain Images

Yeung¹,

Aliasi²,

Haak³

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

freehand ultrasound is the mainstay in prenatal care and fetal growth monitoring. The task of matching corresponding cross-sectional planes in the 3D anatomy for a given 2D ultrasound brain scan is essential in freehand scanning, but challenging. We propose AdLocUI, a framework that Adaptively Localizes 2D Ultrasound Images in the 3D anatomical atlas without using any external tracking sensor. We first train a convolutional neural network with 2D slices sampled from co-aligned 3D ultrasound volumes to predict their locations in the 3D anatomical atlas. Next, we fine-tune it with 2D freehand ultrasound images using a novel unsupervised cycle consistency, which utilizes the fact that the overall displacement of a sequence of images in the 3D anatomical atlas is equal to the displacement from the first image to the last in that sequence. We demonstrate that AdLocUI can adapt to three different ultrasound datasets, acquired with different machines and protocols, and achieves significantly better localization accuracy than the baselines. AdLocUI can be used for sensorless 2D freehand ultrasound guidance by the bedside. The source code is available at https://github.com/pakheiyeung/AdLocUI.

show abstract

Section: Volume-sampled Imagesmentioning

confidence: 54%

Adaptive 3D Localization of 2D Freehand Ultrasound Brain Images

Yeung¹,

Aliasi²,

Haak³

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…Another disadvantage is that minimizing MMD on the instance representation has the risk of changing the feature scale, while regression tasks are fragile to feature scaling. Thus, Representation Subspace Distance (RSD) (Chen et al, 2021b) closes the domain shift through orthogonal bases of the representation spaces, which are free from feature scaling.…”

Section: Mk-mmdmentioning

confidence: 99%

Transferability in Deep Learning: A Survey

Jiang¹,

Yang²,

Wang³

et al. 2022

Preprint

View full text Add to dashboard Cite

The success of deep learning algorithms generally depends on large-scale data, while humans appear to have inherent ability of knowledge transfer, by recognizing and applying relevant knowledge from previous learning experiences when encountering and solving unseen tasks. Such an ability to acquire and reuse knowledge is known as transferability in deep learning. It has formed the long-term quest towards making deep learning as data-efficient as human learning, and has been motivating fruitful design of more powerful deep learning algorithms. We present this survey to connect different isolated areas in deep learning with their relation to transferability, and to provide a unified and complete view to investigating transferability through the whole lifecycle of deep learning. The survey elaborates the fundamental goals and challenges in parallel with the core principles and methods, covering recent cornerstones in deep architectures, pre-training, task adaptation and domain adaptation. This highlights unanswered questions on the appropriate objectives for learning transferable knowledge and for adapting the knowledge to new tasks and domains, avoiding catastrophic forgetting and negative transfer. Finally, we implement a benchmark and an open-source library, enabling a fair evaluation of deep learning methods in terms of transferability.

show abstract

“…1) Maximum Mean Discrepancy based methods. : The Maximum Mean Discrepancy (MMD) based [25]- [27] methods used the maximum mean discrepancy to measure and reduce the distance of extracted feature. Tzeng et.al [2] introduce the adaptation layer with an additional domain confuse loss to learn the domain-invariant representation.…”

Section: Related Workmentioning

confidence: 99%

Time-Series Domain Adaptation via Sparse Associative Structure Alignment: Learning Invariance and Variance

Li¹,

Cai²,

Chen³

et al. 2022

Preprint

View full text Add to dashboard Cite

Domain adaptation on time-series data is often encountered in the industry but received limited attention in academia. Most of the existing domain adaptation methods for time-series data borrow the ideas from the existing methods for non-time series data to extract the domain-invariant representation. However, two peculiar difficulties to time-series data have not been solved. 1) It is not a trivial task to model the domain-invariant and complex dependence among different timestamps.2) The domain-variant information is important but how to leverage them is almost underexploited. Fortunately, the stableness of causal structures among different domains inspires us to explore the structures behind the time-series data. Based on this inspiration, we investigate the domain-invariant unweighted sparse associative structures and the domain-variant strengths of the structures. To achieve this, we propose Sparse Associative structure alignment by learning Invariance and Variance (SASA-IV in short), a model that simultaneously aligns the invariant unweighted spare associative structures and considers the variant information for time-series unsupervised domain adaptation. Technologically, we extract the domain-invariant unweighted sparse associative structures with a unidirectional alignment restriction and embed the domain-variant strengths via a welldesigned autoregressive module. Experimental results not only testify that our model yields state-of-the-art performance on three real-world datasets but also provide some insightful discoveries on the knowledge transfer.

show abstract

Interaction representation-based subspace learning for domain adaptation

Cited by 11 publications

References 7 publications

Adaptive 3D Localization of 2D Freehand Ultrasound Brain Images

Adaptive 3D Localization of 2D Freehand Ultrasound Brain Images

Transferability in Deep Learning: A Survey

Time-Series Domain Adaptation via Sparse Associative Structure Alignment: Learning Invariance and Variance

Contact Info

Product

Resources

About