by cdraptor » Fri Oct 05, 2007 2:59 am
by cdraptor
Fri Oct 05, 2007 2:59 am
I've created some additional videos of the work I am doing with the Speecys SPC-101C. I'm working on getting the developer documents translated to English, but I've already figured out alot. I've updated my C# control application running on Vista to use the System.Speech libraries in .NET 3.0 for creating speech synthesis and voice recognition. I used the speech synthesis to output .wav file, then converted that using sox to .raw format which can be loaded onto the SPC-101C and called within the motion file.
I am working with Speecys on ways in which I load up the .raw file and call them out more realtime - great use would be integrate mail and IM type messaging into the robot. Currently just using the default voice with Vista.
Now here are some examples of using the voice recognition, I have the application provide voice feedback for effect, but I'll be taking that out so the commands execute off quicker. The next video after that is showing the SPC-101C walking around upstairs in my (removed by spam filter) while I am on the laptop controlling it from down in the basement.
I'm going to be looking into the SDK in more detail I see calls which I think will provide direct servo control via TCP/IP socket calls - currently what you are seeing in the application is executing off loaded motion files.
One thing I love about the SPC-101C is that all the data files, minus audio, are ASCII based and the format is pretty easy to figure out, which means tons of opportunities for those to create their own tools.
More details to come - also check out Trossen Robotics forum - I'll be keeping more updated information on there.
I've created some additional videos of the work I am doing with the Speecys SPC-101C. I'm working on getting the developer documents translated to English, but I've already figured out alot. I've updated my C# control application running on Vista to use the System.Speech libraries in .NET 3.0 for creating speech synthesis and voice recognition. I used the speech synthesis to output .wav file, then converted that using sox to .raw format which can be loaded onto the SPC-101C and called within the motion file.
I am working with Speecys on ways in which I load up the .raw file and call them out more realtime - great use would be integrate mail and IM type messaging into the robot. Currently just using the default voice with Vista.
Now here are some examples of using the voice recognition, I have the application provide voice feedback for effect, but I'll be taking that out so the commands execute off quicker. The next video after that is showing the SPC-101C walking around upstairs in my (removed by spam filter) while I am on the laptop controlling it from down in the basement.
I'm going to be looking into the SDK in more detail I see calls which I think will provide direct servo control via TCP/IP socket calls - currently what you are seeing in the application is executing off loaded motion files.
One thing I love about the SPC-101C is that all the data files, minus audio, are ASCII based and the format is pretty easy to figure out, which means tons of opportunities for those to create their own tools.
More details to come - also check out Trossen Robotics forum - I'll be keeping more updated information on there.