Abstract: | 研究期間:10109~10205;A Visual-Based Audio Presentation and Hearing Assistant Device Sound is a very important media for conveying messages. It is indispensable to our life. For example, we can communicate with each other or comprehend the environment by receiving sounds through our ears. Moreover, some sound signals (e.g., car horns) are protective to our safety. However, for people with sensori-neural hearing loss, sounds can barely be heard, and the disease is still incurable in our medical treatment nowadays. Moreover, for other hearing-impaired people, many messages from the sounds are lost. Though some of them can receive sounds through cochlear implants or hearing aids it is not really practical. For example, under the circumstances of strong interfering, the hearing aids would magnify the background voices. This condition would reduce the will of hearing-impaired people to use them. Furthermore, the expensive operation of installing cochlear implants, consumables and the long period needed to adapt to them are daunting to hearing-impaired people. In order to improve the life quality of hearing-impaired people, we begin with three questions: “What troubles do hearing-impaired people experience in daily life?”, “What are the difficulties for them to communicate with other people?” and “Why don’t most of them use cochlear implants or hearing instruments?[28]” To each of the question, we propose several solutions and integrate them into an auditory auxiliary device based on micro-scope monitor. We hope the device can replace the traditional auditory auxiliary device by visional interface. It can provide the function of reading, writing and sound recognition for daily life or communication problems. The device also provides the function of emotion recognition, which helps hearing-impaired people to communicate with hearing people more successfully. The target to be achieved 1. To investigate and develop a technique of background sound recognition which is applicable in embedded systems, and also propose an optimization for programming. 2. To investigate and develop a technique of speech recognition which is applicable in embedded systems. 3.To investigate and develop a technique of emotion recognition which is applicable in embedded systems, and also propose an optimization for programming. 4.To investigate and develop a technique of sound source location detection which is applicable in embedded systems, and also propose an optimization for programming. 5.To investigate and develop the database of sound models for cloud computing which supports users to update the information of new models. 6. The hardware platform of micro-scope monitor includes processor, ADC/DAC, monitor modules, microphone arrays transmission modules of wireless network and RAM modules. 7. According to the real life forms and preference of hearing-impaired people. It is evaluated by questionnaire, pattern selection, and probation test. 8. Design the product not only purely based on functionality, but also the perspective of humanity. The product can increase the emotional connection to real life and satisfy the basic requirements. |