This paper provides a literature review on the topic of deep neural networks and speech recognition. I am providing a review of the scholarly article “deep neural network and speech recognition. The scholarly article is an online book which teaches natural networks and programming paradigm. The book teaches natural networks. It further describes information data programming from which computers are used in the observation of data (Juan Luis Navarro Mesa, Alfonso Ortega, António Te, 2014,pp.2-4). A powerful technique known as deep learning is used during the learning process of the topic neural network (Ronzhin, et al., 2015, pp. 23-25). At the moment, deep learning has been established as one of the very few books that describe the topic of discussion. In addition, the book of neural networks and deep learning has come out as the best solution provider to many problems that deal with image recognition, processing of natural language and speech recognition among others. The online book offers many teachings regarding the main concepts in neural networks and deep learning. The author explains that there is a high likelihood that new learning algorithms promote automatic speech recognition. The first occurrence of the incident happened several decades ago (Laurent Besacier, Adrian-Horia Dediu, Carlos Martí, 2014,pp.27). The algorithm has since made it possible for speech recognition systems to be developed. They are today used to perform various tasks across the world with the use of Gaussian models (Maas, 2016, p. 45).
Section 1.1 represents the broad scan. It is the area where I have talked about how I carried out the research regarding deep neural networks and speech recognition. The broad scan describes how I collected research materials for the topic and how I wrote the bibliography of the online book that had been provided as the area of research. Section 1.2 the area of focus can. It is the area where I have provided a detailed analysis regarding the topic that has been provided. Thereafter, the bibliography was updated to include the additional details that were added in the section. In other words, I can say that filtration was done in this section. Section 1.3 as well as 1.4 provides the literature review. I carried out a thorough review of the online book in this section. Section 1.5 provides an outline of the chapters of the literature review. Finally, introduction of the topic of research is provided in section 1.6.
I carried my research regarding the topic area on the internet. I typed the topic on the Google search engine. I obtained useful links which enabled to complete the research work. Apart from Google, I also researched from VU library and IEEE. I obtained very useful materials which I went ahead and used in my research work.
DATE |
TASK |
ACTION |
COMMENT |
8/5/2017 |
Carried a thorough research regarding the topic |
Identified materials that I would use in completing the assignment |
Noted down main points |
9/5/2017 |
Found important materials that discussed the topic that I had identified. The materials were obtained from different sources, among them Google search, Vu library, and many other scholarly writings |
I spent a good time reading the materials while noting down important points so that I would refer to them later on. |
Created a folder on the desktop and then saved all the all the important materials that I had I identified within the folder. |
10/5/2017 |
Literature reading |
Read five articles regarding the topic |
Identified two very good articles out of the five that I had read. I placed the two articles in a separate folder so that I would use it further. |
11/5/2017 |
Literature reading |
Read other three further materials downloaded from the internet. The materials. |
Established that the materials were irrelevant and hence not in any way related to the research topic. Threw away all the irrelevant materials and concentrated on the materials that had been proved relevant and useful |
12/4/2017 |
Commence the assignment |
Cited the sourced of all the materials first. Immediately followed it with bibliography |
I inserted all the citations as required |
12/04/2017 |
Reading through the paper |
Noted down all the important points. Consulted other people in areas where I had no understanding |
I wrote down the points in a notebook using simple statements which I could easily understand. |
12/5/2017 |
Review of the provided topic |
Read the entire topics that I had obtained from the internet |
Identified the most important areas and then went ahead to carry out a review of the entire document which I was working on. |
13/5/2017 |
Writing of the work |
Wrote the work regarding the provided topic of deep neural networks and speech recognition |
Completed all the parts of the assignment as required |
SOURCE |
KEY USED WORD |
Total number of returned literature |
Total number of collected literature |
Neural networks Models GMM’s and DNN’s |
823 1043 1003 |
2 3 3 |
|
Scholarly sources |
Modelling Dynamic system Coefficient |
5006 4023 3042 |
4 7 4 |
IEEE |
Parameters Speech recognition network |
2400 783 945 |
5 2 6 |
The author states that spoken language has increasingly become a pervasive interface that is used by computing devices. It has become quite difficult for people to understand spoken language since speech signal has to be converted into words before meaningful words can be extracted. Tasks regarding spoken language can be divided into parts which are distinct. The distinct parts should thebe capable of performing functions such as audio signal processing, speech transcription and understanding of natural language.
According to the author, image processing, computer vision and pattern recognition are areas which are closely related and has continuously developed over the years. They have several applications in our day to day lives such as in industry and commerce as well. The growth in the applications has been very rapid and has surpassed even the theoretical advances. As a result, there is need for a handbook for pattern recognition to be developed every five years. The handbook should also provide computer vision.
Source |
Key Used Word |
Total number of returned literature |
Total number of collected literature |
Vu library |
Neural networks Models GMM’s and DNN’s |
823 1043 1003 |
2 3 2 |
Scholarly sources |
Modelling Dynamic system Coefficient |
5006 4023 3042 |
3 3 2 |
IEEE |
Parameters Speech recognition network |
2400 783 945 |
4 1 3 |
Deep neural networks and speech recognition is a relatively tricky topic. The paper explains the relationship that exists between deep neural networks and speech recognition. We try to establish the efficiency of deep neural networks (DNN) during recognition of speech. Several strategies have been employed in ensuring that the best method is identified. Section 1.3.1 provides a clear and elaborate methodology regarding deep neural networks and speech recognition. Approaches are also discussed under the methodology and approaches. The next section, i.e. section 1.3.2 shows methods and framework that were followed during the entire project. The overview of the assignment finally comes in section 1.3.3. Finally, there is section 1.3.5 where there is a discussion. It is in this section where the pros and cons regarding the topic have inclusively been included.
Deep natural network have a close relationship with speech recognition. Take for example a learning algorithm of a new machine. Such machine can promote automatic recognition of speech. Speech recognition systems that we have today across the world have been developed with the aid of EM algorithm. Gaussian mixture models (GMMs) provided the necessary knowledge required for speech recognition. The model created the relationship between acoustic inputs and HMM states (Hinton, et al., 2016, p. 17). Coefficients of Perpetual Linear predictive or Coefficient of Mel Frequency Central are used in the system to represent acoustic input. There are a number of advantages that are associated with GMMs. The advantages of the models enable them to be used for probability distribution modeling. A lot of research has been carried out to constrain GMM so that their valuation speed can be increased. In addition, the valuation has been useful in optimizing trade-off and the available training data. This has ensured that cases of over-fitting have been completely eliminated. GMMs are regarded as very successful models that cannot be replaced by any other. However, it should be noted that GMMs has its fair of disadvantages (IGI Global, Management Association, Information Resources). They lack the necessary statistical information required for modeling processes. Speech is always produced when a small number of parameters are modulated together in a dynamic system. The implication of the case is that underlying structures are dimensionally much lower compared windows that frequently have multiple coefficients. Due to the concern, I believe that there are other models that would perform better than GMMs during acoustic modeling. There are many models that are more efficient in the exploitation of information contained within large windows (K. Sreenivasa Rao, Manjunath K E, 2017,pp16).
Artificially made neural networks combined with error derivatives have the capacity to produce even better models. The resultant models are instrumental in the training of deep neural networks. Deep neural networks have several non-linear and hidden layers. In addition, the DNNs have relatively large output layers (Daniel Graupe, 2016,pp.5-6). The layers are large so that they are capable of accommodating the many states of HMM that arise whenever a different “triphone” models a phone. Very many tied states may still arise even when the states of the HMMs are combined together (Alberto Abad, Alfonso Ortega, António Teixeira, Ca, 2016,pp.19). There are people who have carried out some research and have established that GMMs is inferior to DMMs when it comes to modeling for speech recognition which involves different databases with evenly large vocabularies as well as datasets. A deep neural network has been identified by many people as being the best system model in speech recognition when the data sets are relatively large (Alessandro E.P. Villa, Paolo Masulli, Antonio J. P, 2016,pp.36-38).
The research was to be done at the individual level. That would ensure that there is active participation among all the people who are concerned with the topic. Several persuasive technologies were used. Among the persuasive technologies are customization, simulation, praise, feedback, suggestion, and self-monitoring as well (Simone Bassis, Anna Esposito, Francesco Carlo Mora, 2015,pp.33-37).
A storyboard was used during the research work. Storyboard ensures that there are no cases of bias in the outcomes that are expected. In addition to the efficiency of a storyboard, it provides a very comfortable environment such that all the people who are involved in the research work can express themselves and be understood.
This paper has discussed analyzed and finally discussed the topic of deep neural networks and speech recognition. I used a storyboard to seek the opinion of other people in relation to the topic. However, the people whom I approached were incapable of providing us with the necessary assistance. I therefore resorted to the used of other sources that included internet sources. The sources are Vu library, scholarly sources, and IEEE.
In carrying out the assignment, I used a framework that had a total of two dimensions that include proficiency of operation and technology. The proficiency and technology discussed the various networks alongside their strengths and weaknesses.
It is necessary that we must believe in what we do and also put a lot of effort in them in order to ensure that we become successful. Positivity must, therefore, be encouraged at all times. It has been established that there is a very close relationship between deep natural network and speech recognition. Neural networks perform very important duties, particularly in models. It is necessary that speech must be recognized during modeling processes. Artificially made neural networks when combined with error derivatives, they produce better models (Savchen, 2016,pp 43-44). The resultant models play very important roles in the training of deep neural networks. Deep neural networks have several non-linear and hidden layers.
In addition, the DNNs have relatively large output layers. The layers are large so that they are capable of accommodating the many states of HMM that arise in every new phone model. Very many tied states may still arise even when the states of the HMMs are combined together (United States. Department of Energy,2016,pp.13). There are people who have carried out some research and have established that GMMs is inferior to DMMs when it comes to modeling for speech recognition which involves different databases with evenly large vocabularies as well as datasets. Many people have shown love for the deep neural network. It is, therefore, necessary to network to be developed further so that it can be of great benefit to very many people (inyu Li, Li Deng, Reinhold Haeb-Umbach, Yifan Gong, 2015,pp.9).
The technology of deep neural networks and speech recognition is very important and efforts should be made so that its development can be supported even further. Many neural networks lack the capacity to recognize speech. As a result, it has been impossible to incorporate such networks in the development of models. However, the deep neural network has a great speech recognition. The characteristic has made the network to stand out above the rest.
My topic area is deep neural networks and speech recognition. Deep neural networks and speech recognition is a large topic. In addition, the topical area has not been effectively exhausted in terms of research. The technology is still new and hence many people should come on board so that a lot of research can be done in the area. On my side, I did a lot of research from a series of sources so that I could be able to complete the assignment as had been provided in the instructions. The paper analyses the deep neural network and its role in speech recognition. There are many other neural networks that would also be discussed and looked into. Most systems that have the capacity to recognise speech uses HMMs in an effort to distinguish the variations that exist in speech. However deep neural network is special in a way in the manner in which it recognises speech. The paper, therefore, would tend to pay a close a tension on speech recognition of different neural actions.
References list
Abad, A., Ortega, A. & Teixeira, A., 2016. Advances in Speech and Language Technologies for Iberian Languages: Third International Conference, IberSPEECH 2016, Lisbon, Portugal, November 23-25, 2016, Proceedings. s.l.:Springer.
Arunachalam, S., 2016. Approximate Neural Networks for Speech Applications in Resource-constrained Environments, Chicago: Keyword searching.
Bassis, S., Esposito, A. & Mora, F. C., 2015. Advances in Neural Networks: Computational and Theoretical Issues. s.l.:Springer.
Besacier, L., Dediu, A.-H. & Martí, C., 2014. Statistical Language and Speech Processing: Second International Conference, SLSP 2014, Grenoble, France, October 14-16, 2014, Proceedings. s.l.:Springer.
Chen, C. H., 2015. Handbook of Pattern Recognition and Computer Vision. London, UK: World Scientific.
Dong Yu, Li Deng, 2014. Automatic Speech Recognition: A Deep Learning Approach. s.l.:Springer.
Graupe, D., 2016. Deep Learning Neural Networks: Design and Case Studies. s.l.:World Scientific Publishing Co Inc.
Habernal, I. & Matousek, V., 2013. Text, Speech, and Dialogue: 16th International Conference, TSD 2013, Pilsen, Czech Republic, September 1-5, 2013, Proceedings. New Jersey: Springer.
Hinton, G., 2016. Deep Neural Networks for Acoustic Modeling, s.l.: s.n.
IGI Global,, Management Association, Information Resources. Psychology and Mental Health: Concepts, Methodologies, Tools, and Applications: Concepts, Methodologies, Tools, and Applications. 2013: IGI Global.
inyu Li, L. D., Haeb-Umbach, R. & Gong, Y., 2015. Robust Automatic Speech Recognition: A Bridge to Practical Applications. s.l.:Elsevier Science.
Kurmas Sarma; Muismita Sarme, 2016. Phoneme-Based Speech Segmentation using Hybrid Soft Computing Framework. s.l.:Springer.
Long Cheng, Qingshan Liu, Andrey Ronzhin, 2016. Advances in Neural Networks – ISNN 2016: 13th International Symposium on Neural Networks, ISNN 2016, St. Petersburg, Russia, July 6-8, 2016, Proceedings. s.l.:Springer.
Maas, A. L., 2016. Deep Neural Networks in Speech Recognition. 1 ed. New York: Wiley.
Mesa, J. L. N. & Ortega, A., 2014. Advances in Speech and Language Technologies for Iberian Languages: IberSPEECH 2014 Conference, Las Palmas de Gran Canaria, Spain, November 19-21, 2014, Proceedings. s.l.:Springer.
Rao, S. & Manjunath, K. E., 2017. Speech Recognition Using Articulatory and Excitation Source Features. s.l.:Springer.
Ronzhin, A., Potapova, R. & Németh, G., 2015. Speech and Computer: 18th International Conference, SPECOM 2016, Budapest, Hungary, August 23-27, 2016, Proceedings. New York: Springer.
Savchen, A. V., 2016. Search Techniques in Intelligent Classification Systems. s.l.:Springer.
Tan, T., Li, X., Chen, X. & Zhou, J., 2016. Pattern Recognition: 7th Chinese Conference, CCPR 2016, Chengdu, China, November 5-7, 2016, Proceedings, Part 2. London: Springer.
United States Department of Energy, 2016. Analysis of Pre-Trained Deep Neural Networks for Large-Vocabulary Automatic Speech Recognition. s.l.:United States Department of Energy.
Villa, A. E., Masulli, P. & Antonio, J. P., 2016. Artificial Neural Networks and Machine Learning – ICANN 2016: 25th International Conference on Artificial Neural Networks, Barcelona, Spain, September 6-9, 2016, Proceedings, Part 2. s.l.:Springer.
Železný, M., Habernal, I. & Ronzhin, A., 2013. Speech and Computer: 15th International Conference, SPECOM 2013, September 1-5, 2013, Pilsen, Czech Republic, Proceedings. New York: Springer.
Abad, A., Ortega, A. & Teixeira, A., 2016. Advances in Speech and Language Technologies for Iberian Languages: Third International Conference, IberSPEECH 2016, Lisbon, Portugal, November 23-25, 2016, Proceedings. s.l.:Springer.
Arunachalam, S., 2016. Approximate Neural Networks for Speech Applications in Resource-constrained Environments, Chicago: Keyword searching.
Bassis, S., Esposito, A. & Mora, F. C., 2015. Advances in Neural Networks: Computational and Theoretical Issues. s.l.:Springer.
Besacier, L., Dediu, A.-H. & Martí, C., 2014. Statistical Language and Speech Processing: Second International Conference, SLSP 2014, Grenoble, France, October 14-16, 2014, Proceedings. s.l.:Springer.
Chen, C. H., 2015. Handbook of Pattern Recognition and Computer Vision. London, UK: World Scientific.
Dong Yu, Li Deng, 2014. Automatic Speech Recognition: A Deep Learning Approach. s.l.:Springer.
Graupe, D., 2016. Deep Learning Neural Networks: Design and Case Studies. s.l.:World Scientific Publishing Co Inc.
Habernal, I. & Matousek, V., 2013. Text, Speech, and Dialogue: 16th International Conference, TSD 2013, Pilsen, Czech Republic, September 1-5, 2013, Proceedings. New Jersey: Springer.
Hinton, G., 2016. Deep Neural Networks for Acoustic Modeling, s.l.: s.n.
IGI Global,, Management Association, Information Resources. Psychology and Mental Health: Concepts, Methodologies, Tools, and Applications: Concepts, Methodologies, Tools, and Applications. 2013: IGI Global.
inyu Li, L. D., Haeb-Umbach, R. & Gong, Y., 2015. Robust Automatic Speech Recognition: A Bridge to Practical Applications. s.l.:Elsevier Science.
Kurmas Sarma; Muismita Sarme, 2016. Phoneme-Based Speech Segmentation using Hybrid Soft Computing Framework. s.l.:Springer.
Long Cheng, Qingshan Liu, Andrey Ronzhin, 2016. Advances in Neural Networks – ISNN 2016: 13th International Symposium on Neural Networks, ISNN 2016, St. Petersburg, Russia, July 6-8, 2016, Proceedings. s.l.:Springer.
Maas, A. L., 2016. Deep Neural Networks in Speech Recognition. 1 ed. New York: Wiley.
Mesa, J. L. N. & Ortega, A., 2014. Advances in Speech and Language Technologies for Iberian Languages: IberSPEECH 2014 Conference, Las Palmas de Gran Canaria, Spain, November 19-21, 2014, Proceedings. s.l.:Springer.
Rao, S. & Manjunath, K. E., 2017. Speech Recognition Using Articulatory and Excitation Source Features. s.l.:Springer.
Ronzhin, A., Potapova, R. & Németh, G., 2015. Speech and Computer: 18th International Conference, SPECOM 2016, Budapest, Hungary, August 23-27, 2016, Proceedings. New York: Springer.
Savchen, A. V., 2016. Search Techniques in Intelligent Classification Systems. s.l.:Springer.
Tan, T., Li, X., Chen, X. & Zhou, J., 2016. Pattern Recognition: 7th Chinese Conference, CCPR 2016, Chengdu, China, November 5-7, 2016, Proceedings, Part 2. London: Springer.
United States Department of Energy, 2016. Analysis of Pre-Trained Deep Neural Networks for Large-Vocabulary Automatic Speech Recognition. s.l.:United States Department of Energy.
Villa, A. E., Masulli, P. & Antonio, J. P., 2016. Artificial Neural Networks and Machine Learning – ICANN 2016: 25th International Conference on Artificial Neural Networks, Barcelona, Spain, September 6-9, 2016, Proceedings, Part 2. s.l.:Springer.
Železný, M., Habernal, I. & Ronzhin, A., 2013. Speech and Computer: 15th International Conference, SPECOM 2013, September 1-5, 2013, Pilsen, Czech Republic, Proceedings. New York: Springer.
Abad, A., Ortega, A. & Teixeira, A., 2016. Advances in Speech and Language Technologies for Iberian Languages: Third International Conference, IberSPEECH 2016, Lisbon, Portugal, November 23-25, 2016, Proceedings. s.l.:Springer.
Arunachalam, S., 2016. Approximate Neural Networks for Speech Applications in Resource-constrained Environments, Chicago: Keyword searching.
Bassis, S., Esposito, A. & Mora, F. C., 2015. Advances in Neural Networks: Computational and Theoretical Issues. s.l.:Springer.
Besacier, L., Dediu, A.-H. & Martí, C., 2014. Statistical Language and Speech Processing: Second International Conference, SLSP 2014, Grenoble, France, October 14-16, 2014, Proceedings. s.l.:Springer.
Chen, C. H., 2015. Handbook of Pattern Recognition and Computer Vision. London, UK: World Scientific.
Dong Yu, Li Deng, 2014. Automatic Speech Recognition: A Deep Learning Approach. s.l.:Springer.
Graupe, D., 2016. Deep Learning Neural Networks: Design and Case Studies. s.l.:World Scientific Publishing Co Inc.
Habernal, I. & Matousek, V., 2013. Text, Speech, and Dialogue: 16th International Conference, TSD 2013, Pilsen, Czech Republic, September 1-5, 2013, Proceedings. New Jersey: Springer.
Hinton, G., 2016. Deep Neural Networks for Acoustic Modeling, s.l.: s.n.
IGI Global,, Management Association, Information Resources. Psychology and Mental Health: Concepts, Methodologies, Tools, and Applications: Concepts, Methodologies, Tools, and Applications. 2013: IGI Global.
inyu Li, L. D., Haeb-Umbach, R. & Gong, Y., 2015. Robust Automatic Speech Recognition: A Bridge to Practical Applications. s.l.:Elsevier Science.
Kurmas Sarma; Muismita Sarme, 2016. Phoneme-Based Speech Segmentation using Hybrid Soft Computing Framework. s.l.:Springer.
Long Cheng, Qingshan Liu, Andrey Ronzhin, 2016. Advances in Neural Networks – ISNN 2016: 13th International Symposium on Neural Networks, ISNN 2016, St. Petersburg, Russia, July 6-8, 2016, Proceedings. s.l.:Springer.
Maas, A. L., 2016. Deep Neural Networks in Speech Recognition. 1 ed. New York: Wiley.
Mesa, J. L. N. & Ortega, A., 2014. Advances in Speech and Language Technologies for Iberian Languages: IberSPEECH 2014 Conference, Las Palmas de Gran Canaria, Spain, November 19-21, 2014, Proceedings. s.l.:Springer.
Rao, S. & Manjunath, K. E., 2017. Speech Recognition Using Articulatory and Excitation Source Features. s.l.:Springer.
Ronzhin, A., Potapova, R. & Németh, G., 2015. Speech and Computer: 18th International Conference, SPECOM 2016, Budapest, Hungary, August 23-27, 2016, Proceedings. New York: Springer.
Savchen, A. V., 2016. Search Techniques in Intelligent Classification Systems. s.l.:Springer.
Tan, T., Li, X., Chen, X. & Zhou, J., 2016. Pattern Recognition: 7th Chinese Conference, CCPR 2016, Chengdu, China, November 5-7, 2016, Proceedings, Part 2. London: Springer.
United States Department of Energy, 2016. Analysis of Pre-Trained Deep Neural Networks for Large-Vocabulary Automatic Speech Recognition. s.l.:United States Department of Energy.
Villa, A. E., Masulli, P. & Antonio, J. P., 2016. Artificial Neural Networks and Machine Learning – ICANN 2016: 25th International Conference on Artificial Neural Networks, Barcelona, Spain, September 6-9, 2016, Proceedings, Part 2. s.l.:Springer.
Železný, M., Habernal, I. & Ronzhin, A., 2013. Speech and Computer: 15th International Conference, SPECOM 2013, September 1-5, 2013, Pilsen, Czech Republic, Proceedings. New York: Springer.
Essay Writing Service Features
Our Experience
No matter how complex your assignment is, we can find the right professional for your specific task. Contact Essay is an essay writing company that hires only the smartest minds to help you with your projects. Our expertise allows us to provide students with high-quality academic writing, editing & proofreading services.Free Features
Free revision policy
$10Free bibliography & reference
$8Free title page
$8Free formatting
$8How Our Essay Writing Service Works
First, you will need to complete an order form. It's not difficult but, in case there is anything you find not to be clear, you may always call us so that we can guide you through it. On the order form, you will need to include some basic information concerning your order: subject, topic, number of pages, etc. We also encourage our clients to upload any relevant information or sources that will help.
Complete the order formOnce we have all the information and instructions that we need, we select the most suitable writer for your assignment. While everything seems to be clear, the writer, who has complete knowledge of the subject, may need clarification from you. It is at that point that you would receive a call or email from us.
Writer’s assignmentAs soon as the writer has finished, it will be delivered both to the website and to your email address so that you will not miss it. If your deadline is close at hand, we will place a call to you to make sure that you receive the paper on time.
Completing the order and download