Scalogram Neural Network Activations with Machine Learning for Domestic Multi-channel Audio Classification

Copiaco, Abigail; Ritz, Christian; Fasciani, Stefano; Abdulaziz, Nidhal

doi:10.1109/isspit47144.2019.9001814

Cited by 12 publications

(12 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Because deep learning based on scalograms has been used to successfully classify other types of complex biometric signals, we investigated whether a similar approach can be used to automate HTR detection. According to recent studies, images containing time-frequency data can be very effectively analyzed using a multistage approach where features extracted using an off-the-shelf deep CNN are used to train a support vector machine (SVM) or another type of classifier [33][34][35][36][37] . Combining a pretrained deep CNN with an SVM often outperforms other architectures 33,34,[38][39][40] .…”

mentioning

confidence: 99%

“…The 50-layer ResNet model is a deep CNN that is trained to classify 1,000 image types and includes residual learning, which improves performance by increasing the depth of representation 41 . There is considerable evidence that pretrained CNNs can be used to extract deep features from scalograms and other types of images that are distant from the original datasets used for training 33,34,[38][39][40]42 . Features were extracted from a fully connected layer of ResNet-50 because deeper network layers generate a rich semantic image representation 42 and are well suited for image recognition tasks 34,37,39 .…”

mentioning

confidence: 99%

“…There is considerable evidence that pretrained CNNs can be used to extract deep features from scalograms and other types of images that are distant from the original datasets used for training 33,34,[38][39][40]42 . Features were extracted from a fully connected layer of ResNet-50 because deeper network layers generate a rich semantic image representation 42 and are well suited for image recognition tasks 34,37,39 . Analysis of magnetometer recordings containing >10,000 HTR induced by hallucinogens confirmed that deep feature extraction combined with an SVM can be used to detect the behavior in a very sensitive manner, including low-amplitude and long-duration events.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Automated detection of the head-twitch response using wavelet scalograms and a deep convolutional neural network

Halberstadt

2020

Sci Rep

View full text Add to dashboard Cite

Hallucinogens induce the head-twitch response (HTR), a rapid reciprocal head movement, in mice. Although head twitches are usually identified by direct observation, they can also be assessed using a head-mounted magnet and a magnetometer. procedures have been developed to automate the analysis of magnetometer recordings by detecting events that match the frequency, duration, and amplitude of the HTR. However, there is considerable variability in the features of head twitches, and behaviors such as jumping have similar characteristics, reducing the reliability of these methods. We have developed an automated method that can detect head twitches unambiguously, without relying on features in the amplitude-time domain. To detect the behavior, events are transformed into a visual representation in the time-frequency domain (a scalogram), deep features are extracted using the pretrained convolutional neural network (CNN) ResNet-50, and then the images are classified using a Support Vector Machine (SVM) algorithm. These procedures were used to analyze recordings from 237 mice containing 11,312 HTR. After transformation to scalograms, the multistage CNN-SVM approach detected 11,244 (99.4%) of the HTR. The procedures were insensitive to other behaviors, including jumping and seizures. Deep learning based on scalograms can be used to automate HtR detection with robust sensitivity and reliability. Serotonergic hallucinogens such as psilocybin, d-lysergic acid diethylamide (LSD), and mescaline can induce profound alterations of consciousness, including changes in thought, perception, and mood. The psychedelic effects produced by these drugs are thought to be primarily mediated by activation of the 5-HT 2A receptor 1-4. In recent years, studies have explored the potential therapeutic effects of serotonergic hallucinogens in disorders such as anxiety and depression 5-8. Given the potential therapeutic efficacy of these compounds, there is an increasing need for preclinical behavioral models that can be used to study the pharmacology and mechanism of action of hallucinogens. Administration of LSD and other hallucinogens to mice induces the head-twitch response (HTR), a paroxysmal side-to-side head rotation, which is mediated via 5-HT 2A receptor activation 9,10. The HTR assay is a popular behavioral model for assessing 5-HT 2A activation by hallucinogens in rodents 11-15. Although the non-hallucinogenic LSD analog lisuride is active in some preclinical behavioral models used to study hallucinogens, lisuride does not induce the HTR in mice 16,17. In addition, there is a robust correlation between the potency of hallucinogens in the HTR paradigm and their activity in humans 18. The HTR has traditionally been assessed by direct observation, making data collection time-consuming and subjective. To increase the reliability and throughput of HTR studies, the behavior can be detected using a head-mounted magnet and a magnetometer coil 17. The head movement made during the HTR is highly rhythmic and has a higher frequency than most oth...

show abstract

mentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Automated detection of the head-twitch response using wavelet scalograms and a deep convolutional neural network

Halberstadt

2020

Sci Rep

View full text Add to dashboard Cite

show abstract

“…Scalogram is the modulus of the multiscale wavelet transform, which elucidate time-frequency visualization. The spectro-temporal nature of scalograms makes them desirable for neural networks due to the signal's mapping properties with minimalistic information loss [25]. Scalogram gives the insight into the frequency, energy in time which shown as a function ( , ) in Eq-6.…”

Section: B Scalogrammentioning

confidence: 99%

Classification of Indian Classical Music With Time-Series Matching Deep Learning Approach

et al. 2021

View full text Add to dashboard Cite

Music is a heavenly way of expressing feelings about the world. The language of music has vast diversity. For centuries, people have indulged in debates to stratisfy between Western and Indian Classical Music. But through this paper, an understanding can be fabricated while differentiating the types of Indian Classical Music. Classical music is one of the essential characteristics of Indian Cultural Heritage. Indian Classical Music is divided into two major parts, i.e. Hindustani and Carnatic. Models have been sculptured and trained to classify between Hindustani and Carnatic Music. In this paper, two approaches are used to implement classification models. MFCCs are used as features and implemented models like DNN (1 Layer, 2 Layers, 3 Layers), CNN (1 Layer, 2 Layers, 3 Layers), RNN-LSTM, SVM (Sigmoid, Polynomial & Gaussian Kernel) as one approach. A 3 channels input is created by merging features like MFCC, Spectrogram and Scalogram and implemented models like VGG-16, CNN (1 Layer, 2 Layers, 3 Layers), ResNet-50 as another approach. 3 Layered CNN and RNN-LSTM model performed best among all the approaches.

show abstract

“…Esto es posible luego de obtener una imagen con información multidimencional de la señal procesada. Los espectrogramas (tiempo -frecuencia -amplitud) y escalogramas obtenidos a partir de la transformada de ondícula Wavelet Transform (WT) (tiempo -banda escalada de frecuencia -amplitud) son ejemplos de estas aplicaciones (Zeng et al, 2019;Copiaco et al, 2019).…”

Section: Dominio Tiempo -Frecuenciaunclassified

Exploración de las condiciones del pavimento a partir de la señal acústica producida por un vehículo circulando

Romero¹

View full text Add to dashboard Cite

Hace ya casi cuatro años que inició este viaje académico, motivado por entender un poco más sobre un entorno que nunca deja de vibrar y sonar.Durante este proceso, siempre he contado con el apoyo incondicional y lleno de amor de Tatiana y Valentina. Gracias por estar conmigo, llenando mis días de vida y alegría, a pesar de lo apremiante que pudo llegar a ser esta travesía. Este logro es también suyo.Mi gratitud eterna es para mis Padres y Hermanos. Pese a que nos separan varios meridianos, esta etapa también es fruto de su tesón, guía, abnegación y entusiasmo.A toda mi familia, especialmente a mi primo Oscar. Su conanza también hizo posible que este camino se pudiera iniciar.Un agradecimiento especial a mi director, César Asensio. Ésta investigación ha requerido de él su paciencia e imprescindible dedicación.Quiero agradecer también al personal investigador y administrativo del Grupo de Investigación i2a2 por su asistencia y soporte, sin los cuales la consecución del presente trabajo habría visto jornadas muy complicadas. Gracias Guillermo, Ignacio y Eva. A los compañeros del Grupo i2a2: Pablo, Ricardo, Luis, Yolanda, Andrés, Daniel y Juan Manuel. Gracias por sus enseñanzas y su colaboración.Debo agradecer particularmente a Luis Sigcha, compañero desde las aulas del pregrado. Su don de gente, amistad y conocimiento han sido sin duda un regalo muy preciado.A los compañeros de Labenac por su soporte en las tareas instrumentales para la realización de varios experimentos.También agradezco a Mariana, Pepe, Nelson y Naty, quienes en más de una ocasión me brindaron el calor de su hogar con amabilidad y bondad.A los compañeros becarios, quienes han echo de esta estancia una experiencia más familiar, especialmente a Fredy, Zamir y Giovanny.

show abstract

Scalogram Neural Network Activations with Machine Learning for Domestic Multi-channel Audio Classification

Cited by 12 publications

References 11 publications

Automated detection of the head-twitch response using wavelet scalograms and a deep convolutional neural network

Automated detection of the head-twitch response using wavelet scalograms and a deep convolutional neural network

Classification of Indian Classical Music With Time-Series Matching Deep Learning Approach

Exploración de las condiciones del pavimento a partir de la señal acústica producida por un vehículo circulando

Contact Info

Product

Resources

About