Speech recognition is a way of enabling a computer to decode human voice in more precise and perfect way. The computer translates the analog waves of voice into digital by analyzing the sound. The voice analysis is crucial for security reasons which help security engineers safeguard the nation against terrorist and curbing other dangerous crimes. An important reason why speech recognition is a hard problem is noise. Speech can uttered in an environment of a car hooting, radio sounds, and computer sounds among other things all this result to noise. In speech recognition an individual must identify a specific sound that come from a particular signal to recognize the sound effectively. Echo effect is another kind of sound, where the speech signal bounces on the surrounding object and reaches the microphone a few seconds later.
We Can Help Write Your Essay on Speech Recognition Now!
Another reason why speech recognition is a hard problem is the signs of a body language. A person’s speaker apart from communicating with speech also does with body signals such as hand waving, postures, eye movements, among other related things. ASR misses this information. This issue aired within the research place multimodality, where outcome conducted on how to integrate body language to help the human-computer communication.
Continuous speech is another reason why speech recognition is hard. The statement does not have natural pauses between boundaries of a word; the interludes appear on the syntactic level, such as after a sentence which brings a difficult for speech recognition. After the stage of recognition into phones categories, they grouped into words (Simonyan et al. pp. 44). Nevertheless, we disregard word boundary uncertainty; this is a difficult problem. One strategy to simplify the whole process is to give the pauses between famous words (James, pp.33). It works for short command communication, but as the possible length of utterances increases, breaks get inefficient and cumbersome.
One feature I would like to have an artificial element is the ability to record the sounds effectively. The synthetic speech recognizers should in a position to keep all the noise in details whether from loud noise or moderate of sound. Another feature I would want the synthetic speech to have is the ability to identify the origin of the sound, for instance, to be able to identify the specific object (James, pp. 33). Also, another feature I would want for artificial is the ability to distinguish the fake sound and the original sound for instance if a person is screaming awkwardly.
Some of the problems of human speech that are solved are how to separate words from a background that is noisy and overcrowded for instance when a person shout at you in the street being able to separate the words and sound correctively (James, pp. 33). Another issue of human speech that resolved is identifying words especially when someone chatters (Simonyan et al. pp. 44). The beetle makes squeaking sounds which are like hammer which are notable in speech recognition. The carriage makes a noisy sound that is recognizable from a far distance especially the sound of horses sunning. A speaker can create a pumping sound where electromagnet attracts and repels by a magnet, the sound is easily recognized. The beaker also produces an intense sound which is recognizable if there is not interference from other background sounds.
In conclusion, human speech recognition is a way of enabling a computer to decode human voice. In Speech Recognition a person must perform activities such as filtering and identifying specific noise from certain signal so as to recognize the speech. Echo effect. Speech does not have natural pauses between the boundaries of the word. Another feature I would want the synthetic speech to have is the ability to identify the origin of the sound, for instance, to be able to identify the specific object.
Works cited
Simonyan, Kristina, et al. “New developments in understanding the complexity of human speech production.” Journal of Neuroscience 36.45 (2016): 11440-11448.
James, Alex Pappachen. “Heart rate monitoring using human speech spectral features.” Human-centric Computing and Information Sciences 5.1 (2015): 33.
EffectivePapers.com is a professional essay writing service committed to writing non-plagiarized custom essays, research papers, dissertations, and other assignments of top quality. All academic papers are written from scratch by highly qualified essay writers. Just proceed with your order, and we will find the best academic writer for you!