DGR: Gender Recognition of Human Speech Using One-Dimensional Conventional Neural Network

Alkhawaldeh, Rami S.

doi:10.1155/2019/7213717

Cited by 56 publications

(33 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…To develop such a model for classification problems, one must choose the right feature from the human voice by which both gender and region of a person can be determined easily. There are most common features like MFCCs, Mel-scaled power spectrogram (Mel), power spectrogram chroma (Chroma), spectral contrast (Contrast), and tonal centroid features (Tonnetz) employed for gender recognition using 1D CNN model (Alkhawaldeh, 2019). Fundamental frequency, Energy, Spectral flatness, entropy, intensity, zero-crossing rate, etc.…”

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

Gender and region detection from human voice using the three-layer feature extraction method with 1D CNN

Uddin

Pathan

Hossain

et al. 2021

Journal of Information and Telecommunication

View full text Add to dashboard Cite

Analysing the human voice has always been a challenge to the engineering society for various purposes such as product review, emotional state detection, developing AI, and much more. Two basic grounds of voice or speech analysis are to detect human gender and the geographical region based on accent. This study presents a three-layer feature extraction method from the raw human voice to detect the gender as male or female, as well as the region from where that voice belongs. Fundamental frequency, spectral entropy, spectral flatness, and mode frequency have been calculated in the first layer of feature extraction. On the other hand, Mel Frequency Cepstral Coefficient has been used to extract the features in the second layer and linear predictive coding in the third layer. Regular voice contains some noises which have been removed with multiple audio data filtering processes to get noise-free smooth data. Multi-Outputbased 1D Convolutional Neural Network has been used to recognize gender and region from a combined dataset which consists of TIMIT, RAVDESS, and BGC datasets. The model has successfully predicted the gender with 93.01% and region with 97.07% accuracy. This method works better than usual state-ofthe-art methods in separate datasets along with the combined dataset on both gender and region classification.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

“…To solve this noise problem needs a pre-processing process for feature extraction. Pre-emphasis, frame blocking, the hamming window are used in (Alkhawaldeh, 2019;Yusnita et al, 2017) as a pre-processing phase for removal of the noise.…”

Section: Literature Reviewmentioning

confidence: 99%

Gender and region detection from human voice using the three-layer feature extraction method with 1D CNN

Uddin

Pathan

Hossain

et al. 2021

Journal of Information and Telecommunication

View full text Add to dashboard Cite

show abstract

“…Apparently, not only pitch-related features, but also Mel-Frequency Cepstrum Coefficients (MFCC) and energy-related features can play an important role for gender identification. A combination among the above-mentioned features also ensured particularly better results, as demonstrated in [21], [22]. Some other approaches were also tried to identify the gender of a speaker directly from raw audio signals like in [23] where a CNN was trained to directly extract features from the raw signal in a filtering stage and then the classification was performed.…”

Section: Related Workmentioning

confidence: 99%

Gender Identification in a Two-Level Hierarchical Speech Emotion Recognition System for an Italian Social Robot

Guerrieri¹,

Braccili²,

Sgrò³

et al. 2021

Preprint

View full text Add to dashboard Cite

The real challenge in Human Robot Interaction (HRI) is to build machines capable of perceiving human emotions so that robots can interact with humans in a proper manner. It is well known from the literature that emotion varies accordingly to many factors. Among these, gender represents one of the most influencing one, and so an appropriate gender-dependent emotion recognition system is recommended. In this paper, a two-level hierarchical Speech Emotion Recognition (SER) system is proposed: the first level is represented by the Gender Recognition (GR) module for the speaker’s gender identification; the second is a gender-specific SER block. Specifically for this work, the attention was focused on the optimisation of the first level of the proposed architecture. The system was designed to be installed on social robots for hospitalised and living at home elderly patients monitoring. Hence, the importance of reducing the software computational effort of the architecture also minimizing the hardware bulkiness, in order for the system to be suitable for social robots. The algorithm was executed on the Raspberry Pi hardware. For the training, the Italian emotional database EMOVO was used. Results show a GR accuracy value of 97.8%, comparable with the ones found in literature.

show abstract

“…They claimed that the model is very effective even if the amount of learning data is limited. Alkhawaldeh et al [23] described a gender classification model with the help of one dimension convolution neural network. They used the features, such as Mel Spectrogram, Mel Frequency Cepstral Coefficients, as the single dimension input sequence to CNN for training of the model.…”

Section: Literature Surveymentioning

confidence: 99%

A rough set theory and deep learning based predictive system for gender recognition using audio speech

Yasmin

Das

Nayak

et al. 2021

Preprint

View full text Add to dashboard Cite

Speech is one of the most delicate medium through which gender of the speakers can easily be identified. Though the related research has shown very good progress in machine learning but recently, deep learning has imparted a very good research area to explore the deficiency of gender discrimination using traditional machine learning techniques. In deep learning techniques, the speech features are automatically generated by the reinforcement learning from the raw data which have more discriminating power than the human generated features. But in some practical situations like gender recognition, it is observed that combination of both types of features sometimes provides comparatively better performance. In the proposed work, we have initially extracted and selected some informative and precise acoustic features relevant to gender recognition using entropy based information theory and Rough Set Theory (RST). Next, the audio speech signals are directly fed into the deep neural network model consists of Convolution Neural Network (CNN) and Gated Recurrent Unit network (GRUN) for extracting features useful for gender recognition. The RST selects precise and informative features, CNN extracts the locally encoded important features, and GRUN reduces the vanishing gradient and exploding gradient problems. Finally, a hybrid gender recognition system is developed combining both generated feature vectors. The developed model has been tested with five bench mark and a simulated dataset to evaluate its performance and it is observed that combined feature vector provides more effective gender recognition system specially when transgender is considered as a gender type together with male and female.

show abstract

DGR: Gender Recognition of Human Speech Using One-Dimensional Conventional Neural Network

Cited by 56 publications

References 17 publications

Gender and region detection from human voice using the three-layer feature extraction method with 1D CNN

Gender and region detection from human voice using the three-layer feature extraction method with 1D CNN

Gender Identification in a Two-Level Hierarchical Speech Emotion Recognition System for an Italian Social Robot

A rough set theory and deep learning based predictive system for gender recognition using audio speech

Contact Info

Product

Resources

About