Global ETD Search

Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.

Results

Showing 1 to 15 of 15 for “"raw waveform"”.

  1. Interpreting intermediate feature representations of raw-waveform deep CNNs by sonification

    … works that address the interpretability of raw waveform based deep neural networks (DNNs) for audio processing focus on interpreting spectral and frequency response information, often limiting to visual and signal theoretic means of interpretation, solely for the first layer. This work …

    glasgow Repository record for Interpreting intermediate feature representations of raw-waveform deep CNNs by sonification (opens in a new tab) Download PDF for Interpreting intermediate feature representations of raw-waveform deep CNNs by sonification (opens in a new tab)

  2. Leveraging Foundation Models for Wave-Based Sensing Classification Tasks

    … different feature representations such as the raw waveform and time-frequency feature representations. Additionally, trainable encoders to improve the feature embeddings and classifcation performance are also explored in this work. The results show that the use of AVES as a foundation model is …

    tamu Repository record for Leveraging Foundation Models for Wave-Based Sensing Classification Tasks (opens in a new tab) Embargoed until 2028-05-01

  3. Leveraging Foundation Models for Wave-Based Sensing Classification Tasks

    … different feature representations such as the raw waveform and time-frequency feature representations. Additionally, trainable encoders to improve the feature embeddings and classifcation performance are also explored in this work. The results show that the use of AVES as a foundation model is …

    tamu Repository record for Leveraging Foundation Models for Wave-Based Sensing Classification Tasks (opens in a new tab) Embargoed until 2028-05-01

  4. Leveraging Foundation Models for Wave-Based Sensing Classification Tasks

    … different feature representations such as the raw waveform and time-frequency feature representations. Additionally, trainable encoders to improve the feature embeddings and classifcation performance are also explored in this work. The results show that the use of AVES as a foundation model is …

    tamu Repository record for Leveraging Foundation Models for Wave-Based Sensing Classification Tasks (opens in a new tab) Embargoed until 2028-05-01

  5. End-to-End Music Transcription Using Fine-Tuned Variable-Q Filterbanks

    … available, the viability of learning from the raw waveform of recordings widens. Deep neural networks have been shown to perform feature extraction and classification jointly. With sufficient data, optimal filters which operate in the time-domain may be learned in place of conventional …

    rit Repository record for End-to-End Music Transcription Using Fine-Tuned Variable-Q Filterbanks (opens in a new tab)

  6. Spike-Based Architectures for Energy-Efficient Audio Enhancement

    … spectrogram masking and supplements it with raw waveform-based reconstruction in the time domain. Our experiments show that SNNs -- especially the Dual-Signal model -- are able to achieve competitive denoising performance while substantially lowering the computational cost—opening up …

    rit Repository record for Spike-Based Architectures for Energy-Efficient Audio Enhancement (opens in a new tab)

  7. Neural waveform generation for source-filter vocoding in speech synthesis

    … has been the introduction of neural network waveform synthesis models, which take the role of a vocoder in a traditional speech synthesis system. However, a gap remains both in understanding and computational efficiency of the algorithms between the recent raw waveform neural vocoders and the …

    aalto Repository record for Neural waveform generation for source-filter vocoding in speech synthesis (opens in a new tab)

  8. A Percussion Method to Classify Bolt Torque for Decommissioned Wind Turbine Blade (DWTB) Structures in the Field

    … Wavelet Transform (CWT), MFCC-image, and raw waveform inputs for a four-convolution layer Convolutional Neural Network (CNN). Results show that CNNs trained on mel-spectrogram images consistently achieve the highest performance, with a peak accuracy of 98.3% on independent nut-segregated …

    houston Repository record for A Percussion Method to Classify Bolt Torque for Decommissioned Wind Turbine Blade (DWTB) Structures in the Field (opens in a new tab) Embargoed until 2027-12-01

  9. UNSUPERVISED SEGMENTAL MODELING OF SPEECH FOR LOW RESOURCE APPLICATIONS

    … and extracts frame-level representation from the raw waveform. In the second stage, a differentiable boundary detector finds variable-length segments. In the last stage, the model predicts future segments to learn segment representations. Experiments show that our model outperforms existing phone …

    jhu Repository record for UNSUPERVISED SEGMENTAL MODELING OF SPEECH FOR LOW RESOURCE APPLICATIONS (opens in a new tab)

  10. Spatial audio and spatial audio-visual learning

    … how can classic signal processing based sound waveform feature extraction methods benefit from the powerful expressiveness of deep neural network?</p> <p>2. Is the widely-adopted process of spectrogram extraction which treats sound as ordinary 2D image an optimal representation? If not, can we …

    oxford Repository record for Spatial audio and spatial audio-visual learning (opens in a new tab)

  11. Speech enhancement by using deep learning algorithms

    … further explores a new cascaded architecture on raw waveform input against the complexity of auditory perception.<br/>Chapter 5 brings a new combination method in speech enhancement, contrasting mapping-based and masking-based methods, and proposing a parallel dual-module system, the Compensation …

    soton Download PDF for Speech enhancement by using deep learning algorithms (opens in a new tab)

  12. 심전도 기반 관상동맥 죽상경화증 탐지를 위한 인공신경망 개발

    … labeled with the corresponding CACS. Only the raw waveform of the ECG was used as input for ​​model training. Among the 12 leads, the 3 leads closest to the orthogonal axis (leads I, II, V2) were used as input. Moreover, the 10-second ECGs were divided into 2.5 second intervals for data …

    ajou Repository record for 심전도 기반 관상동맥 죽상경화증 탐지를 위한 인공신경망 개발 (opens in a new tab) Embargoed until 9999-12-31

  13. Interpretable speaker diarisation for child language environment analysis

    … The other network receives the input speech as raw waveforms in its time-domain representation. Interpreting these models requires that the explanations convey that the model has latched on to the idea of gender discriminative information, such as a difference in pitch and formant frequencies. …

    nwu-za Repository record for Interpretable speaker diarisation for child language environment analysis (opens in a new tab)

  14. Harnessing airborne and spaceborne LiDAR for large-area assessment of canopy structural complexity

    … retrieval to high-altitude and future spaceborne waveform LiDAR observations. An encoder–decoder transformer was developed to reconstruct high-resolution canopy profiles from waveforms, acquired by NASA’s Land Vegetation and Ice Sensor (LVIS) at the Smithsonian Environmental Research Center, using …

    rit Repository record for Harnessing airborne and spaceborne LiDAR for large-area assessment of canopy structural complexity (opens in a new tab)

  15. Multimodal Neuron Classification based on Morphology and Electrophysiology

    Categorizing neurons into different types to understand neural circuits and ultimately brain function is a major challenge in neuroscience. While electrical properties are critical in defining a neuron, its morphology is equally important. Advancements in single-cell analysis methods have allowed …

    wvu Repository record for Multimodal Neuron Classification based on Morphology and Electrophysiology (opens in a new tab)