Deep Hashing Network for Unsupervised Domain Adaptation

Venkateswara, Hemanth; Eusébio, José; Chakraborty, Sudipta; Panchanathan, Sethuraman

doi:10.1007/978-3-030-45529-3_4

Cited by 32 publications

(6 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Traditional TL benchmarks have more balanced ratios between source and target domains. For example, in the office dataset, the ratios range between 1.87 to 6.59; 42 for the extended office-Home dataset, the ratios are close to 1 43 , and for the Photo-Art-Cartoon-Sketch (PACS) dataset the ratios range from 1.14 to 2.35. 44 In the presented Smart Maintenance Living Lab (SMLL) dataset, the ratio ranges from 8.6 to 21.6, meaning less data is available for each target domain.…”

Section: Transfer Learning In Pdm Tasksmentioning

confidence: 97%

Transfer Learning for Anomaly Detection Using Bearings' Vibration Signals

Nieves-Avendano,

Deschrijver,

Van Hoecke

2023

The International Journal of Acoustics and Vibration

View full text Add to dashboard Cite

Predictive maintenance is becoming increasingly important in the industry. Despite considerable advances in data collection and data-driven models, there are still limitations when deploying models in practice. One of the main limitations is the large datasets required to train these models. As a potential solution, transfer learning can be used to reuse knowledge acquired from large datasets for similar tasks under different conditions. This paper investigates the transferability of the specific anomaly detection scenario, an unsupervised learning task with heavily imbalanced label distribution. This paper uses a dataset generated from a bearing test platform in which bearings are run until failure under different operating conditions. A lightweight deep learning model, MobileNetV2, is employed to create a baseline model capable of detecting anomalies for a specific working condition. The model is then adapted using transfer learning to identify anomalies under new operating conditions with limited data accurately. The results show that the data for new conditions is insufficient to train an adequate model and that transfer learning can overcome this limitation. The adapted models can detect anomalies before the expert's knowledge reference value. Although this shows that transfer learning can detect anomalies earlier, the results must be evaluated carefully to avoid false positives. While anomaly detection aims to identify changes in feature distributions, transfer learning aims to align different feature distributions. Transfer learning for unsupervised learning has rarely been explored. To the best of our knowledge, this is one of the few works addressing it in the context of predictive maintenance for anomaly detection.

show abstract

Section: Transfer Learning In Pdm Tasksmentioning

confidence: 97%

Transfer Learning for Anomaly Detection Using Bearings' Vibration Signals

Nieves-Avendano,

Deschrijver,

Van Hoecke

2023

The International Journal of Acoustics and Vibration

View full text Add to dashboard Cite

show abstract

“…With relevant insights from [57], [58], we extend our hashingbased SC approach for domain adaptation that includes unsupervised domain adaptation between the sender's and receiver's knowledge base, semantic extraction at the sender, and unsupervised hashing for the receiver. We employ Multi-Kernel Maximum Mean Discrepancy (MKMMD) [58] to quantify the distribution difference between sender and receiver datasets in a reproducing-kernel Hilbert space which help reducing knowledge base disparity through nonlinear data alignment. In particular, with the extended framework, we seek to minimize (slight abuse of notations to distinguish sender and receiver and extending the ideas from [58])…”

Section: B With Domain Adaptive Hashing (Dah)mentioning

confidence: 99%

“…We employ Multi-Kernel Maximum Mean Discrepancy (MKMMD) [58] to quantify the distribution difference between sender and receiver datasets in a reproducing-kernel Hilbert space which help reducing knowledge base disparity through nonlinear data alignment. In particular, with the extended framework, we seek to minimize (slight abuse of notations to distinguish sender and receiver and extending the ideas from [58])…”

Section: B With Domain Adaptive Hashing (Dah)mentioning

confidence: 99%

See 1 more Smart Citation

Understand-Before-Talk (UBT): A Semantic Communication Approach to 6G Networks

Pokhrel

Choi

2023

IEEE Trans. Veh. Technol.

View full text Add to dashboard Cite

In Shannon theory, semantic aspects of communi-• Technical (Level A): How correctly can the bits/packets be transmitted from the sender to the receiver? • Semantics (Level B): How accurately do the bits/packets carry the required meaning from the sender to the receiver? • Effectiveness (Level C): How well does the meaning transmitted (conveyed) influence the intended course of action at the receiver? At level A, semantic aspects of communication are assumed to be irrelevant to the technical engineering problems. Shannon noted "Frequently the messages have meaning; that is they refer to or are correlated according to some system with certain physical or conceptual entities. These semantic aspects of communication are irrelevant to the engineering problem" [1]. This is one of the fundamental assumptions of Shannon's theory. The relaxation of this fundamental assumption for communication between a sender and a receiver can jointly consider all three levels of problem, which can be referred to

show abstract

“…Combined with a new distance loss named maximum density divergence, Zhang et al. [9] propose an adversarial tight match model to improve the ability of the domain adaptation by adversarial training and metric learning.…”

Section: Introductionmentioning

confidence: 99%

Adversarial discriminative domain adaptation for modulation classification based on Ulam stability

Ren

Chen²,

Yang

2023

IET Radar Sonar & Navi

View full text Add to dashboard Cite

Domain adaptation modulation classification aims to improve the cross‐domain robustness of modulation classification, with the basic idea of mitigating the domain shift between the label‐rich source domain and label‐poor target domain in the latent common feature subspace. Due to the complex and heterogeneous wireless propagation conditions, how to construct the optimal latent common feature subspace is a hard problem. In this letter, the authors introduce Ulam stability to character the boundary of the latent common feature space. These theorems can provide guidance for cross‐domain feature extraction and eliminate the domain shift. Based on the proposed stable theorems, the authors define the Ulam loss and present a new method named Adversarial Discriminative Domain Adaptation with Ulam stability (U‐ADDA). Experiments on public datasets prove the effectiveness and generality of the method.

show abstract

Deep Hashing Network for Unsupervised Domain Adaptation

Cited by 32 publications

References 31 publications

Transfer Learning for Anomaly Detection Using Bearings' Vibration Signals

Transfer Learning for Anomaly Detection Using Bearings' Vibration Signals

Understand-Before-Talk (UBT): A Semantic Communication Approach to 6G Networks

Adversarial discriminative domain adaptation for modulation classification based on Ulam stability

Contact Info

Product

Resources

About