Lanczos kernel based spectrogram image features for sound classification

Özer, İlyas; Ozer, Zeynep; Fındık, Oğuz

doi:10.1016/j.procs.2017.06.020

Cited by 12 publications

(7 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Greco et al [60] Environmental sound and acoustic scene classification Sound/event credit network [45,47,112] Sound event detection Adaboost Ykhlef et al [174] Sound classification Using the restricted boltzmann machine, DNN was trained Ozer et al [124] Fig. 5 Block diagram of the signal preprocessing stage [119,147] parameters and frequency sub-bands using the WPT were derived from the sub-bands to describe the distribution of wavelet coefficients.…”

Section: Transformersmentioning

confidence: 99%

A review on lung disease recognition by acoustic signal analysis with deep learning networks

2023

View full text Add to dashboard Cite

Recently, assistive explanations for difficulties in the health check area have been made viable thanks in considerable portion to technologies like deep learning and machine learning. Using auditory analysis and medical imaging, they also increase the predictive accuracy for prompt and early disease detection. Medical professionals are thankful for such technological support since it helps them manage further patients because of the shortage of skilled human resources. In addition to serious illnesses like lung cancer and respiratory diseases, the plurality of breathing difficulties is gradually rising and endangering society. Because early prediction and immediate treatment are crucial for respiratory disorders, chest X-rays and respiratory sound audio are proving to be quite helpful together. Compared to related review studies on lung disease classification/detection using deep learning algorithms, only two review studies based on signal analysis for lung disease diagnosis have been conducted in 2011 and 2018. This work provides a review of lung disease recognition with acoustic signal analysis with deep learning networks. We anticipate that physicians and researchers working with sound-signal-based machine learning will find this material beneficial.

show abstract

Section: Transformersmentioning

confidence: 99%

A review on lung disease recognition by acoustic signal analysis with deep learning networks

2023

View full text Add to dashboard Cite

show abstract

“…To apply an image classification approach, audio data were pre-processed and transformed into image data. For example, features extracted from the Spectrogram image were shown to improve the performance of acoustic event classifications [11]. Besides, Spectrogram images were also being used for rapid speaker recognition and artificial speech detection [12].…”

Section: Figure 1 the Illustration Of Speaker Identification Versus Speaker Verificationmentioning

confidence: 99%

“…In this work, rather than using conventional signal processing to extract features from speech, we proposed to use data transformation to apply the techniques from other domains such as image processing for artificial speech detection. The proposed approach is motivated by the success of deploying image classification techniques to sounds classification and speaker recognition [11]. An ensemble classifier in the form of random forest (RF) was used to generate the artificial speech model.…”

Section: Figure 1 the Illustration Of Speaker Identification Versus Speaker Verificationmentioning

confidence: 99%

Artificial speech detection using image-based features and random forest classifier

Tan

Hijazi

Kok³

et al. 2022

IJ-AI

View full text Add to dashboard Cite

The ASVspoof 2015 Challenge was one of the efforts of the research community in the field of speech processing to foster the development of generalized countermeasures against spoofing attacks. However, most countermeasures submitted to the ASVspoof 2015 Challenge failed to detect the S10 attack effectively, the only attack that was generated using the waveform concatenation approach. Hence, more informative features are needed to detect previously unseen spoofing attacks. This paper presents an approach that uses data transformation techniques to engineer image-based features together with random forest classifier to detect artificial speech. The objectives are two-fold: (i) to extract image-based features from the melfrequency cepstral coefficients representation of the speech signal and (ii) to compare the performance of using the extracted features and Random Forest to determine the authenticity of voices with the existing approaches. An audio-to-image transformation technique was used to engineer new features in classifying genuine and spoof voices. An experiment was conducted to find the appropriate combination of the engineered features and classifier. Experimental results showed that the proposed approach was able to detect speech synthesis and voice conversion attacks effectively, with an equal error rate of 0.10% and accuracy of 99.93%.

show abstract

“…Bu, veri kümelerinde derin öğrenme modellerinin sınıflandırma performansını kısıtlamıştır. Bu sorunun çözümü, önceden eğitilmiş olarak bilinen bir yöntemle gelmiştir [44][46] [48]. Bu yöntemin arkasındaki fikir, metnin genel temsilini öğrenmek ve çok miktarda etiketlenmemiş metinden oluşan genişletilmiş derlemi eğitmektir.…”

Section: Kelime Temsil Yöntemiunclassified

Twitter Platformundan Elde Edilen Türkçe Saldırgan Dil Derlemi

YILMAZ

Özer

Gökçen³

2022

Mühendislik Bilimleri Ve Araştırmaları Dergisi

View full text Add to dashboard Cite

Sosyal medya platformlarında kullanıcıların paylaşımlar arasında saldırgan dil barındıran içeriklerin önemli oranda arttığı gözlemlenmiştir. Çalışma Türkçe dilinde bu sorunun çözümüne katkı sağlamayı amaçlamaktadır. Bu çalışmada Twitter platformundan elde edilen bir veri seti oluşturulmuştur. 14752 Türkçe tweet metninden oluşan bu veri seti etiketleyiciler tarafından manuel olarak etiketlenmiştir. Buna ek olarak oluşturulan veri seti kullanılarak LSTM (Long ShortTerm Memory) ve GRU (Gated Recurrent Units) modellerinin sınıflandırma performansları karşılaştırılmıştır. Çalışmada ikili ve çoklu sınıflandırma yapılmıştır. Saldırgan dil ile ilgili Türkçe için çoklu sınıflandırma yapılan ilk çalışmadır. Bunlara ek olarak Twitter platformundan 1 milyon 860 bin tweet metninden oluşan genişletilmiş derlem elde edilmiştir. Burada word2vec yöntemi ile kelime temsilleri elde edilmiştir. Böylelikle genişletilmiş derlem kullanımının sınıflandırma performanslarına katkısı karşılaştırılmıştır. Çalışmada yapılan ikili sınıflandırma da genişletilmiş derlem kullanımıyla en yüksek performans GRU modeli F1-skor değeri %94,49’dur. Bu sebeple çoklu sınıflandırma yapılırken GRU modeli kullanılmıştır. Çoklu sınıflandırmada elde edilen sınıflandırma performans değerleri genişletilmiş derlemin katkısıyla GRU F1-makro değeri %71,97 ve %54,10’dur. Bu alanda Türk dili literatürüne katkı sağlamak amacıyla mevcut çalışmanın veri setleri ve genişletilmiş derlem kelime vektörleri paylaşılacaktır.

show abstract

Lanczos kernel based spectrogram image features for sound classification

Cited by 12 publications

References 15 publications

A review on lung disease recognition by acoustic signal analysis with deep learning networks

A review on lung disease recognition by acoustic signal analysis with deep learning networks

Artificial speech detection using image-based features and random forest classifier

Twitter Platformundan Elde Edilen Türkçe Saldırgan Dil Derlemi

Contact Info

Product

Resources

About