AWERProcedia Information Technology & Computer Science Vol 04 (2013) 764-769 3 rd World Conference on Innovation and Computer Sciences 2013 Turkish Speech Recognizing Voice Control of Television Nursel Yalçın *, Computer Education Department, Gazi University. Muhammed Mutlu Yapıcı, Computer Engineering, Gazi University. Suggested Citation: Yalçın N. & Yapıcı M., M.Turkish Speech Recognizing Voice Control of Television. AWERProcedia Information Technology & Computer Science. [Online]. 2013, 04, pp 764-769. Available from: www.awercenter.org/pitcs Received December 04, 2012; revised January 18, 2013; accepted March 04, 2013. Selection and peer review under responsibility of Prof. Dr. Fahrettin Sadıkoglu, Near East University. 2013 Academic World Education & Research Center. All rights reserved. Abstract Speech recognition is the most popular technological development easing the life. Although the works on speech recognition in Turkish is quite limited, it is possible to control computers, cars, security systems etc. and even a house can be controlled trough voice without even needing any other controlling device. In this work, it is aimed to develop sample software responding the voice commands to control television without using the remote controller. Keywords: Speech Recognition, Voice Control of Television, Voice Control; *ADDRESS FOR CORRESPONDANCE: Nursel Yalçın, Computer Education Department, Gazi University, E-mail Address: nurselyalcin@gmail.com
1. Introduction With the proliferation of services of computers, the number of computer users increase day by day. As a result of these developments, the ways have been sought to ensure faster and efficient communication. Voice recognition technology is also one of the disciplines that shape these searches. Speech recognition is recognization of the human voice by the computer. Speech recognition allows computer users to use computers independently of the hardware elements as keyboard and mouse. This also increases the interest on speech recognition technologies. Speech recognition is currently used in many fields as telephone banking, telephone shopping, forensic applications, training. Undoubtedly, computers and televisions are two of the communication tools that the people spend a lot of time together. The time spent with a man's television and computer was calculated almost 3 hours and 12 minutes per a day. And this time is increasing every year. In our country, rate of the watching television has been identified as 94% [1]. Given this situation, the demand for devices and software, which is prepared for television technologies, is expected to be undeniable proportions. However, as a result of the research, there are no studies to control the TV with the sound in Turkey. Controlling the TV with transceiver devices was also observed in abroad, however it brings an extra cost to the user. Obligation to keep this device instead of remote control devices was found to be a disadvantage. Televisions can be controlled using the human voice without using remote control by the developed software. The only disadvantage of the software is that we need a keyboard and key protocol for every different device. The only disadvantage of the software is that we need a keyboard and key protocol for every different device. So, any software is not working for every device. Therefore, before the developing the software protocol of the device and the key codes must be known. The developed software costs are vanishingly small. The computer allows us to control the TV from anywhere in any environment. If this system is developed, televisions and any devices will be controlled without the remote control. It is undoubtedly a great and important development for today's technology. In the Second chapter of the article, program procedures and stages of the preparation of the program will be focused on. 2. Developing the software Controlling the television that is the most frequently used communication tool of modern man, by voice commands without the remote control is aimed with the developed software and the circuit. Thus, the elimination of defects of the remote control device is intended. In the preparation of voice control system, the information about X3 hiremco satellite device, which will be controlled by voice commands, was collected (Figure 1). The satellite protocol and key codes were determined by a receiver that was inserted into a computer's serial port and WinLIRC software. Figure 1-Hiremco X3 satellite receiver The most commonly used remote control protocols are SIRC, RC5, SHARP, NEC, Nokia, ITT. X3 Hiremco receiver uses the NEC protocol. NEC protocol structure is shown in Figure 2. 765
Figure 2- NEC protocol structure NEC protocol sends 32-bit packet information. After the parity bit, following a 16-bit key code sends the opposite of the key code for verification purpose. The waiting times necessary for obtaining the information about 32-bit data is described in Figure 3 for "1" key [2]. Figure 3- The waiting times for 1 key For example, according to the waiting times in the first line in the Figure 3, derivation of the bits is as follows: 624 1607 611 506 1 0 When the bits is achieved for all waiting times in Figure 3, for "1" key, 01FD value (device code) +92 (key code) +6C (opposite of key code) is obtained at 16 number system. serialport1.write(new byte[] { 01, 253, 146, 109 }, 0, 4); Signal sent from the computer is set to 38 khz frequency which is detected by a satellite device via an electronic circuit which was developed with using pic programming. Pic 16F628A, IR LED, 4-MHz crystal, resistor and a battery was used for developing the electronic circuit. The electronic circuit is shown on Figure 4. Figure 4- Frequency tuning circuit 766
2.1. User interface designing and software developing After the preparation of the necessary circuit for the remote control, the software has been prepared. Visual Studio 2005 C#.NET and Dikte.Api Demo Software were used to developing the software. Twenty commands were described on the program. The most commonly used television functions were taken into account to determine these Twenty commands. The commands that were described on the program and the operations corresponding to the commands are as follows: Commands The operations One Open Channel 1 Two Open Channel 2 Three Open Channel 3 Four Open Channel 4 Five Open Channel 5 Six Open Channel 6 Seven Open Channel 7 Eight Open Channel 8 Nine Open Channel 9 Zero Open Channel end of the 0 as 10 Menu Open The Menu Escape Escape From Menu Next Channel Open Next Channel Previous Channel Open Previous Channel OK Okey Open Open the TV Clase Close the TV After determination of the commands, interface of the program was designed with C# objects as shown in Figure 5. The flow chart of program is shown on Figure 6. Figure 5- Software interface 767
Start Is there any command? N Stop Y N Is the command valid? Send data to circuit according to command Send data to 38 khz satellite receiver Operate the command Figure 6- The flow chart of program In the software, the command issued by the user is perceived by the system. If the command is one of the described commands, Television is controlled by performing operation of the command. For example, when command "one" has been sent, operation of the command is runned and channel 1 is opened. Screenshot of this operation is shown on Figure 7. 768
Figure 7- Screenshot of the command one Because of the command of "one" was described in the system, pairing procedure has been realized. As a result, the command's data is transferred to the electronic circuit via computer. And then, the circuit sends the data with the 38 KHz frequencies to the satellite device. Thus, channel 1 is opening (Figure 8). 3. Conclusions and advices Figure 8- screen of the TV when command of one has been sent Today, mass communication has an important role admittedly quite. Although computers are one of the most widely used communication tools in our country and around the world, television usage rate is undeniable high. Effective and efficient use of these most requested devices is so important. Voice recognition technology has an important role in human-computer communication. The software and devices, which has been developed via voice recognition technology, emerges in many areas. The television control software with voice commands has also been developed by using speech recognition for televisions that acquired an important place in our daily lives. Being independent of users provides a great advantage for the software. The software which is developed with the support of Turkish speech recognition Is of great importance because of it is the first time in Turkey. With increasing the performance of the software, requesting by many television users will be inevitable. Time that people spend to watch TV in Turkey show that, this is possible. References İstatistikler.net(2010)Türkiye'de Okuma ve İzleme Oranları, <http://istatistikler.net/izleme.html> (2010, Nisan 30) Ak, Y. (2004) Nec Protokolü, Elektronikçinin El Kitabı, 1-10. 769