Skip to content
Free planNo credit card requiredStart writing
AI for Research home
AI for Research home
Dashboard
Sign inGet started

Search 250 million research papers

Read the abstract, see who wrote it and who cites it, and get the citation in APA, MLA, Chicago or Harvard.

About 10,000 papers for “Spectrogram”

Underwater Acoustic Classification Based on Spectrogram Transformer

Dan Zou, Lina Lu, Jizhou Guo et al. · 2023

Underwater acoustic classification is an important method for underwater target detection. In the past decade, with the rapid development in the theory and technology of deep learning, it has been applied to various audio classification scenarios, such as speech recognition and acoustic scene classification. However, due to the complexity and variability of underwater environment, underwater acoustic classification is always a challenging task. This paper presents an attention-based model for underwater acoustic classification based on spectrogram transformer. By training the model with the public underwater acoustic dataset DeepShip, it achieves a result of 89.2% accuracy, outperforming the traditional classification approaches.

Cited by 2

Continuous Arabic Speech Segmentation using FFT Spectrogram

M.M. Awais, Waqas Ahmad, S. Masud et al. · 2006

This paper describes a phoneme segmentation algorithm that uses Fast Fourier Transform (FFT) spectrogram. The algorithm has been implemented and tested for utterances of continuous Arabic speech of 10 male speakers that contain almost 2346 phonemes in total. The recognition system determines the phoneme boundaries and identifies them as pauses, vowels and consonants. The system uses intensity and phoneme duration for separating pauses from consonants. Intensity in particular is used to detect two specific consonants (/r/, /h/) when they are not detected through the spectrographic information. Segmentation accuracy of 95.39% for the overall system has been achieved.

Cited by 18

Spectrogram features for audio and speech analysis

Ian McLoughlin, Lam Pham, Yan Song et al. · arXiv (Cornell University) · 2026

Spectrogram-based representations have grown to dominate the feature space for deep learning audio analysis systems, and are often adopted for speech analysis also. Initially, the primary motivator for spectrogram-based representations was their ability to present sound as a two dimensional signal in the time-frequency plane, which not only provides an interpretable physical basis for analysing sound, but also unlocks the use of a wide range of machine learning techniques such as convolutional neural networks, that had been developed for image processing. A spectrogram is a matrix characterised by the resolution and span of its two dimensions, as well as by the representation and scaling of each element. Many possibilities for these three characteristics have been explored by researchers across numerous application areas, with different settings showing affinity for various tasks. This paper reviews the use of spectrogram-based representations and surveys the state-of-the-art to question how front-end feature representation choice allies with back-end classifier architecture for different tasks.

Cited by 0Open access

Dynamical Spectrogram, An Aid For The Deaf

A. A. Soltani Farani, E.H.S. Chilton, Robin Shirley · 2007

Visual perception of speech through spectrogram reading has long been a subject of research, as an aid for the deaf or hearing impaired. Attributing the lack of success in this type of visual aids mainly to the static form of information presented by the spectrograms, this paper proposes a system of dynamic visualisation for speech sounds. This system samples a high resolved, auditorybased spectrogram, with a window of 20 milliseconds duration, so that exploiting the periodicity of the input sound, it produces a phase-locked sequence of images. This sequence is then animated at a rate of 50 images per second to produce a movie-like image displaying both the time-varying and time-independent information of the underlying sound. Results of several preliminary experiments for evaluation of the potential usefulness of the system for the deaf, undertaken by normal-hearing subjects, support the quick learning and persistence of the gestures for small sets of single words and motivate further...

Cited by 0

Low-cost spectrogram based counterfeit medicine detection

Amit Kumar Mishra, Mohamed Hoosain Essop · arXiv (Cornell University) · 2019

Contaminated substances such as counterfeit medication and food contami-nated with pesticide residue is a pandemic of utmost urgency. Spectroscopy and chromatography methods are often used but are expensive and complex and as such a need exists for a device that can be easily operated in developing commu-nities. We present a hacked visible spectrometer based contaminated substance detector using machine learning. The Support Vector Machine (SVM), Logistic Regression, linear Regression and Convolutional Neural Network (CNN) models have been implemented and are trained on the acquired spectrum data. Our results show that a lowcost method of identifying contaminated substances is achievable with very high accuracy.

Cited by 1Open access

Acoustic Scene Classification with Spectrogram Processing Strategies

Helin Wang, Yuexian Zou, Dading Chong · arXiv (Cornell University) · 2020

Recently, convolutional neural networks (CNN) have achieved the state-of-the-art performance in acoustic scene classification (ASC) task. The audio data is often transformed into two-dimensional spectrogram representations, which are then fed to the neural networks. In this paper, we study the problem of efficiently taking advantage of different spectrogram representations through discriminative processing strategies. There are two main contributions. The first contribution is exploring the impact of the combination of multiple spectrogram representations at different stages, which provides a meaningful reference for the effective spectrogram fusion. The second contribution is that the processing strategies in multiple frequency bands and multiple temporal frames are proposed to make fully use of a single spectrogram representation. The proposed spectrogram processing strategies can be easily transferred to any network structures. The experiments are carried out on the DCASE 2020 Task1 datasets, and the results show that our method could achieve the accuracy of 81.8% (official baseline: 54.1%) and 92.1% (official baseline: 87.3%) on the officially provided fold 1 evaluation dataset of Task1A and Task1B, respectively.

Cited by 11Open access

Adaptive spectrogram for time-frequency signal analysis

Xiaobing Sun, Zheng Bao · 2002

A novel method for the determination of the window parameters of adaptive spectrograms is given in this paper. It is based on detecting the maximum value and the width of the peak of the Radon transform of the modulus of the ambiguity function of the signals. The proposed method effectively reduces the cross-terms and the noise of linear frequency modulated signals compared with the Wigner distribution and the classical fix window spectrogram. A possible further extension of the method is also given to fit it to larger classes of signals.

Cited by 1

Performance of an expert spectrogram reader. II

Alexander I. Rudnicky, Ronald A. Cole, Victor W. Zue · The Journal of the Acoustical Society of America · 1978

Analysis of the videotaped performance of an expert spectrogram reader revealed the control strategy and sources of knowledge used by the expert while reading spectrograms. Spectrograms were read in two stages: segmentation and labeling. During segmentation, which was usually performed in a single pass from left to right, segment markers were placed on the spectrogram to indicate roughly phone-sized units. Segmentation was accomplished through the detection of spectral discontinuities, although the precise placement of segment markers was guided by a detailed knowledge of acoustic phonetics. Labeling was not performed left to fight, and was usually guided by an “easiest segments first” strategy. The vast majority of segments were identified through the recognition of characteristic patterns and salient features. The identification of other segments required the application of knowledge about coarticulation and phonological rules. Several experiments revealed that the expert did not use higher-order knowledge (e.g., lexical, syntactic, or semantic) during labeling—performance was slightly better on nonsense utterances than English sentences. Implications of the expert's performance will be discussed for theories of speech perception, machine recognition of speech, and the development of aids for the speech- and hearing-impaired.

Cited by 9Open access

Signal analysis of heart sound using spectrogram

Wan Nor Zakaria, Wan Muhammad Aizuddin · UiTM Institutional Repositories (Universiti Teknologi MARA) · 2014

The heart sounds of human in normal good condition typically contain two distinct sound of S1 and S2 that occur in each heartbeat cycle. The respective frequency and time latencies of each tone in heartbeat are important parameters to monitor cardiac function and diagnosis for any abnormality of heart functions. A conventional phonocardiogram analysis using stethoscope can be subjective and imprecise because of heart sound classification analysis depends on the physician interpretation. This paper proposes a spectrogram method which is one of digital signal processing technique to analyze heart sound pattern and to classify for its various diseases related.

Cited by 0Open access

Spectrogram Features for Audio and Speech Analysis

Ian McLoughlin, Lam Pham, Yan Song et al. · Applied Sciences · 2026

Spectrogram-based representations have grown to dominate the feature space for deep learning audio analysis systems, and are often adopted for speech analysis also. Initially, the primary motivation behind spectrogram-based representations was their ability to present sound as a two-dimensional signal in the time–frequency plane, which not only provides an interpretable physical basis for analysing sound, but also unlocks the use of a range of machine learning techniques such as convolutional neural networks, which had been developed for image processing. A spectrogram is a matrix characterised by the resolution and span of its dimensions, as well as by the representation and scaling of each element. Many possibilities for these three characteristics have been explored by researchers across numerous application areas, with different settings showing affinity for various tasks. This paper reviews the use of spectrogram-based representations and surveys the state-of-the-art to question how front-end feature representation choice allies with back-end classifier architecture for different tasks.

Cited by 0Open access

PD-VOST: Parkinson’s Disease Voice Spectrogram Transformer

Ilias Tougui, Mehdi Zakroum, Ouassim Karrakchou et al. · 2025

Deep learning (DL) techniques are increasingly used for diagnosing Parkinsons Disease (PD) due to their non-invasive nature and accessibility. This study introduces PD-VOST (Parkinsons Disease Voice Spectrogram Transformer), an innovative DL model that employs the Transformer architecture for PD diagnosis using voice data collected via smartphones. Unlike conventional transfer learning approaches that typically use image-pretrained models for audio tasks, PD-VOST leverages a model pre-trained specifically on audio data. This specialization allows for more effective fine-tuning on voice recordings from individuals with PD and healthy controls. Our model achieved an average AUC of 95.89% and an average AUPRC of 87.11%, outperforming state-of-the-art methods. These results underscore the potential of employing audio-specific pre-trained models in advancing the early detection and management of PD using the voice as a biomarker.

Cited by 0

Using Spectrogram to Evaluate Dysarthric Treatment Effectiveness

Shu-Er Lee · Rehabilitation practice and science · 1990

Dysarthria may be defined as a group of speech disorders resulting from disturbance in muscular control due to damage in the central or peripheral nervous system. It encompasses coexisting neurogenic disorders of several or all the basic processes of speech - respiration, phonation, articulation, resonance and prosody. Although auditory discrimination can provide subjective evaluation of dysarthric speech before and after treatment, objective evaluation can help us to establish standards for clinical evaluation of acoustic charade- ristics.Six dysarthric and four normal control subjects were evaluated in this study. Sound spectrographic analysis was done of speech samples in vowel prolongation /ah/, oral diadochokinetic movement /p ^ t ^ k ^ / and a 6-word sentence. The results show that differential diagnosis of dysarthria can be made from the spectrogram qualitatively, and the efficacy of treatment can be measured quantitatively. The sound spectrogram provides objectively more detailed and complete analysis of dysarthric speech. It is a significant tool for speech pathologist to use in clinical evaluation and research of dysarthric patients.

Cited by 0Open access

Mean rates in time and aggregate spectrogram.

Seth Frey (282788), Robert L. Goldstone (282789) · Figshare · 2015

<p>The left panel shows the mean rate in each group, at each round. The right panel shows a spectrogram (with a window size of 20 rounds) for resampled observed data. These figures show changes in rotation over the sequence of 200 rounds of play. In the spectrogram, the brightness of a pixel indicates the amplitude of the corresponding frequency component. These panels show statistically significant increases in the rate of periodic behavior, in both the time and frequency domains.</p>

Cited by 0Open access

1H Nuclear Magnetic Resonance Parameters and Spectrogram

Huang, Jing · Figshare · 2023

Method: <sup>1</sup>H NMR spectra was used to detect the conformation of the hydrolyzates. The mixture of samples was 500 μl, the experimental group contained 30 μg enzyme, water-soluble substrate with the final concentration of 18 mM, and 1 mM Ca<sup>2+</sup>, Ba<sup>2+</sup>, or Sr<sup>2+</sup>. Assays were run at 40℃ for 10min and 40min. The standard was 5 mg D-(+)-glucose; and the control group included water-soluble substrate without addition of enzyme. Then 100μl D2O was added to each group before NMR spectra. Finally, all the samples were sent to the Institute for Advanced Research in Central South University for testing hydrolyzates by using Bruker 600 MHz spectrometer (AVANCE III 600M, Switzerland). <br> Results: This is the result of <sup>1</sup>H NMR spectroscopy. 1H NMR analysis showed QsGH97a has inverting mechanism to produce β-glucose. The peaks of H-1 α and H-1 β were shown in the group of D-(+)-Glucose, “Ca<sup>2+</sup> 10min” represents that the reaction mixture is measured at Ca<sup>2+</sup> environment for 10min, “Ca<sup>2+</sup> 40min” exhibits that the reaction mixture is detected at Ca<sup>2+</sup> environment for 40min. Ba<sup>2+</sup>-treated groups and Sr<sup>2+</sup>-treated groups are identical to Ca<sup>2+</sup>-treated groups.

Cited by 0Open access