Figure 9 Comparison of two navigation interface based on speak-to-VR VRPN server
Related Figures (14)
Fig. 1 Flow chart for weight vector calculation The microphone array and the distance between the ele- ments of the array are shown in Fig. 4. The speak-to-VR VRPN server processes spoken com- mands and compares the result with the dictionary content (Table 1) after receiving a word as a 1D signal from microphone array. Then, the associated button is activated and sent to the client application through the wireless network. Navigation and object manipulation applications were developed in the client side to evaluate user perfor- mance. Since it is difficult to use only the speak-to-VR interface for both navigation and object manipulation, this interface was used in combination with other HMI. The Table 2 Two dictionaries used in the navigation study max 20 letters) and dictionary 2 (8 words with min and max similar to dictionary 1) as seen in Table 2. Fig. 6 Pair of magic bracelets Fig. 7 Relation between the length of the words and reduction in the processing speed (increasing processing delay) Table 3 Speech processing and total processing time for navigation dictionary 1 in Table 2 Table 4 Speech processing and total processing time for navigation dictionary 2 in table 2 Table 5 Two dictionary used in manipulation study Table 6 Speech processing and total processing time for manipulation dictionary 1 in table 5 extremely important and the words need to be carefully selected. Table 7 Speech processing and total processing time for manipulation dictionary 2 in Table 5 Table 8 Random accuracy test of speak-to-VR VRPN server Fig. 8 The level of user satisfaction using speak-to-VR interface for interaction and navigation result, it is recommended to select shorter words and include fewer words in the dictionary to make it as fast as possible.