First you create the new Java DukeSpeaks project (Apache web server for windows)
First you create the new Java DukeSpeaks project in the usual way. Again, you need to modify the Java Build Path. This time, however, you don t add external JARs but open the Projects page and checkmark the FreeTTS project. This makes the resources of the FreeTTS project available to the DukeSpeaks project as well. This applies, too, to the external JARs that you added to the FreeTTS project. However, these JARs must be marked for export in the FreeTTS project. This is currently not the case. So you must once again edit the Java Build Path of the FreeTTS project. To do so, select the project in the Package Explorer, right-click, and select the Properties context function. In the dialog that appears, select the Java Build Path category. Then open the Order and Export page. There checkmark all FreeTTS JARs, thus making them available to all projects that build on the FreeTTS project. To avoid having the example files from Chapter 1 littering the Package Explorer, you should create a new working set. To do so, follow these steps: 1. Click the drop-down button (down-arrow) on the toolbar of the Package Explorer and choose the Select Working Set function. In the dialog that now appears, press the New button. 2. In the next dialog, select Java as the working set type and press the Next button. 3. Finally, enter dukeSpeaks as the name and checkmark the FreeTTS and DukeSpeaks projects. From now on, the Package Explorer displays only these two projects. By invoking the Deselect Working Set function you can restore the original state. A Short Excursion into Speech Synthesis Before you start extending the FreeTTS system, you should get acquainted with the basics of speech synthesis and with the architecture of the FreeTTS system. Speech synthesis works in several steps: 1. A Tokenizer breaks the text into syntactical units (tokens). In general, these are words and numbers, including the punctuation. 2. Some tokens, such as numbers, are converted into words. 3. A Phraser analyzes the word list and organizes it into phrases (sentences and para-sentences). Phrasing establishes the basis for the later decoration of the speech output with pauses and melody. 4. A Segmenter analyzes the words and with the help of a lexicon assigns a syllable structure to each word. 5. The Pause Generator inserts a pause in front of each phrase. 6. The Intonator analyzes the syllables and assigns an emphasis and a pitch to each syllable. 7. In a further step and depending on the voice used, some phonemes are replaced by others. 8. The duration for each phoneme is determined. 9. The Contour Generator assigns an envelope curve to each syllable. 74 Chapter 5
Get account with us and you will get completely access to our free web templates database with over 10.000 templates in it to build your website.Don’t wait, go and check free web templates.