FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4310872040> ?p ?o ?g. }

• W4310872040 endingPage "110" @default.
• W4310872040 startingPage "96" @default.
• W4310872040 abstract "This article proposes a new source model and training scheme to improve the accuracy and speed of the multichannel variational autoencoder (MVAE) method. The MVAE method is a recently proposed powerful multichannel source separation method. It consists of pretraining a source model represented by a conditional VAE (CVAE) and then estimating separation matrices along with other unknown parameters so that the log-likelihood is non-decreasing given an observed mixture signal. Although the MVAE method has been shown to provide high source separation performance, one drawback is the computational cost of the backpropagation steps in the separation-matrix estimation algorithm. To overcome this drawback, a method called “FastMVAE” was subsequently proposed, which uses an auxiliary classifier VAE (ACVAE) to train the source model. By using the classifier and encoder trained in this way, the optimal parameters of the source model can be inferred efficiently, albeit approximately, in each step of the algorithm. However, the generalization capability of the trained ACVAE source model was not satisfactory, which led to poor performance in situations with unseen data. To improve the generalization capability, this article proposes a new model architecture (called the “ChimeraACVAE” model) and a training scheme based on knowledge distillation. The experimental results revealed that the proposed source model trained with the proposed loss function achieved better source separation performance with less computation time than FastMVAE. We also confirmed that our methods were able to separate 18 sources with a reasonably good accuracy." @default.
• W4310872040 created "2022-12-19" @default.
• W4310872040 creator A5001243214 @default.
• W4310872040 creator A5061981123 @default.
• W4310872040 creator A5075702573 @default.
• W4310872040 date "2023-01-01" @default.
• W4310872040 modified "2023-09-25" @default.
• W4310872040 title "FastMVAE2: On Improving and Accelerating the Fast Variational Autoencoder-Based Source Separation Algorithm for Determined Mixtures" @default.
• W4310872040 cites W1512388774 @default.
• W4310872040 cites W1543386260 @default.
• W4310872040 cites W1552314771 @default.
• W4310872040 cites W1572099565 @default.
• W4310872040 cites W1845880232 @default.
• W4310872040 cites W1887941968 @default.
• W4310872040 cites W2014768838 @default.
• W4310872040 cites W2058341666 @default.
• W4310872040 cites W2067295501 @default.
• W4310872040 cites W2072548008 @default.
• W4310872040 cites W2086286498 @default.
• W4310872040 cites W2096855653 @default.
• W4310872040 cites W2113990625 @default.
• W4310872040 cites W2116064496 @default.
• W4310872040 cites W2117678320 @default.
• W4310872040 cites W2127851351 @default.
• W4310872040 cites W2139302694 @default.
• W4310872040 cites W2165698076 @default.
• W4310872040 cites W2168273590 @default.
• W4310872040 cites W2221409856 @default.
• W4310872040 cites W2408744528 @default.
• W4310872040 cites W2412956798 @default.
• W4310872040 cites W2460742184 @default.
• W4310872040 cites W2734774145 @default.
• W4310872040 cites W2750446090 @default.
• W4310872040 cites W2766672686 @default.
• W4310872040 cites W2792498316 @default.
• W4310872040 cites W2800675406 @default.
• W4310872040 cites W2808718747 @default.
• W4310872040 cites W2889341949 @default.
• W4310872040 cites W2892163332 @default.
• W4310872040 cites W2894785362 @default.
• W4310872040 cites W2901552243 @default.
• W4310872040 cites W2911579794 @default.
• W4310872040 cites W2922004249 @default.
• W4310872040 cites W2936446744 @default.
• W4310872040 cites W2946555236 @default.
• W4310872040 cites W2952218014 @default.
• W4310872040 cites W2954049404 @default.
• W4310872040 cites W2954996726 @default.
• W4310872040 cites W2962715207 @default.
• W4310872040 cites W2962866211 @default.
• W4310872040 cites W2963035245 @default.
• W4310872040 cites W2963375116 @default.
• W4310872040 cites W2963969588 @default.
• W4310872040 cites W2968723745 @default.
• W4310872040 cites W2972460025 @default.
• W4310872040 cites W2979850772 @default.
• W4310872040 cites W2989942677 @default.
• W4310872040 cites W3015788098 @default.
• W4310872040 cites W3094316140 @default.
• W4310872040 cites W3096831136 @default.
• W4310872040 cites W3100968126 @default.
• W4310872040 cites W3106955634 @default.
• W4310872040 cites W3117310166 @default.
• W4310872040 cites W3141797743 @default.
• W4310872040 cites W3151851237 @default.
• W4310872040 cites W3163652268 @default.
• W4310872040 cites W3198425604 @default.
• W4310872040 cites W4206634335 @default.
• W4310872040 doi "https://doi.org/10.1109/taslp.2022.3214763" @default.
• W4310872040 hasPublicationYear "2023" @default.
• W4310872040 type Work @default.
• W4310872040 citedByCount "1" @default.
• W4310872040 countsByYear W43108720402023 @default.
• W4310872040 crossrefType "journal-article" @default.
• W4310872040 hasAuthorship W4310872040A5001243214 @default.
• W4310872040 hasAuthorship W4310872040A5061981123 @default.
• W4310872040 hasAuthorship W4310872040A5075702573 @default.
• W4310872040 hasBestOaLocation W43108720402 @default.
• W4310872040 hasConcept C101738243 @default.
• W4310872040 hasConcept C111919701 @default.
• W4310872040 hasConcept C11413529 @default.
• W4310872040 hasConcept C120317606 @default.
• W4310872040 hasConcept C127162648 @default.
• W4310872040 hasConcept C134306372 @default.
• W4310872040 hasConcept C153180895 @default.
• W4310872040 hasConcept C154945302 @default.
• W4310872040 hasConcept C155032097 @default.
• W4310872040 hasConcept C177148314 @default.
• W4310872040 hasConcept C178790620 @default.
• W4310872040 hasConcept C185592680 @default.
• W4310872040 hasConcept C204030448 @default.
• W4310872040 hasConcept C2776864781 @default.
• W4310872040 hasConcept C31258907 @default.
• W4310872040 hasConcept C33923547 @default.
• W4310872040 hasConcept C41008148 @default.
• W4310872040 hasConcept C43126263 @default.
• W4310872040 hasConcept C45374587 @default.
• W4310872040 hasConcept C50644808 @default.

• next