Talking Signs - more testing

I tried putting the Go Talk Buttons in a muffin tin but that didn't work out. The muffin tins were too deep and the dogs didn't really want to paw the buttons when they were in there. I tried putting several layers of cardboard in the muffin tin to raise the buttons up but that didn't work either.

Then it occurred to me that I should try Velcro. I didn't want to use the Talking Symbols Velcro so I searched the house and finally found some and a board to attach the buttons too. I first tried regular Velcro and it didn't work that well as the buttons still moved too much. The Industrial Strength Velcro was better. Now I had a new problem. If the dogs pawed at the buttons just the right way the battery compartment opened and the top came off. This can be solved by either putting a larger piece of Velcro on the back to hold the battery compartment closed or using a piece of tape.

If only I could get my dogs to put some weight on the button when they pawed it. Right now they are not and the button doesn't always play the message when they paw it. Sometimes what happens is that the light comes on but no message is played. I duplicated this by rapidly pressing the button twice.

Talking Signs arrived

I got my talking signs yesterday. I ended up buying 6 of the Go Talk Buttons and 3 Talking Symbols. The Talking Symbols are twice the price of the Go Talk Buttons per sign.

The sound quality appears to be stronger and better on the Talking Symbols. Both will allow 10 seconds of recording.

Both have a small red LED that lights up when recording and when playing. The Go Talk Button takes one battery and the Talking Symbol three.

The magnet in the Talking Symbols is stronger than the Go Talk Button. The Talking Symbols come with velcro as well.

Just to get the dogs using them I recorded "yes" on both and rewarded them for getting it to play back the recording.

The Talking Symbols are more sensitive as well. My dogs had to often paw several times at the Talking Button before actually pressing the button and making it play the recording. Taari my younger sheltie (13lb and 14" at the withers) has become better at it so it may just be a matter practice. The Buttons move quite a bit and at one point last night Taari was starting to play with it like she does a Babble ball (a ball that plays phrases when moved). It was amusing to see her batting it around between her paws. There is a larger surface to touch on the Talking Symbols (more than 4 times the area of the Button). The Go Talk Button requires more pressure than the Talking Symbol.

It is not readily apparent which one is more robust. The Go Talk Buttons may be more slobber proof given that the Talking Symbols have a grill over the speaker. The speaker appears to be made of metal though. It is my plan to have the dogs paw the talking signs so hopefully slobber will not be an issue.

I think that the Go Talk Buttons will need to be placed in something or attached to a board to hold them still. I will have to think more about this.

The Talking Symbols could be used on their own as is.

Here is a Youtube video showing both kinds of talking signs and Taari playing with them.

More Silent Signs

Today I continued working on getting the dogs to target the correct coaster. At first I started with only the coaster with the picture of the game on it. Then I added a blank coaster. Both dogs were pretty successful in the morning session so in the afternoon session I decided to switch games. They were also successfully pawing the coaster of the other game. Next I added a blank coaster. Sometimes helping them out by pointing to the coaster. In the evening session I decided to have the dogs go to a mat and then I would release them with okay. Taari seemed to be less successful. Then I noticed that she was looking at both coasters and when I released her she frequently went to the one she was looking at. I realized why this was happening as that is what I expect when I do gambling work with her. So I now have to be careful and make sure that she is looking at the right coaster when I release her. She did start to release herself and go to the correct coaster. I am not too worried about her releasing herself because she was picking the right coaster and was not waiting for me say "yes that is the right one".

I haven't added the question yet as I am unsure if I want the question to be "Which game do you want?" or "Which game is this?"

One problem I am having is making sure that they touch the coaster with a paw. This is where I think that the talking signs would help. I would wait until I heard the sign talk (i.e the dog would have to hit the sign and cause the recording to play). There would be feedback to the dog that they had touched the sign appropriately - it would speak the word or words associated with what they touched. I would also not be confusing the dog by clicking it for a bad paw or not noticing when they did hit it. (Aside - I can see that a Chicken Camp would help with my skills.)

Silent Signs

While waiting for the for the talking signs I decided to work on getting Spice and Taari to paw a coaster with a picture on it. I went too fast at first and have back tracked so I will describe what I am now doing.

1. First I started with a "blank" coaster and worked on their pawing technique (clicking and treating for a good paw on the coaster).
   
2. Then I printed a picture of the Dog Tornado Game and put it on a coaster. I put the Tornado Game between my feet and held it so that they couldn't try to play the whole Tornado Game. Then I put the coaster nearby. I pointed to the coaster and when they pawed it I clicked and put a treat in the Tornado Game.
   
3. Eventually I stopped pointing and they were pawing it on their own.

I will repeat 3 and maybe 2 if needed in more sessions. After that I think that the next step is to do the same with the Dog Brick Game.

Keyboards, Signs and Buttons

After hearing about and reading a paper by Alexandre Pongrácz Rossi and César Ades I thought it would be interesting to try teach my dogs to use something similar to the keyboard device that they used. The device in the paper turns out to be quite expensive so I am going to give other devices a try (they are essentially talking signs). I will associate something with the sign that will identify it to the dog and the dog will press the sign/button and a word or short message will be played.

I can think of at least a few applications or experiments I could do with these devices.

1. Teach the dog to press a specific sign/button when it wants something (i.e. go outside, food, play, petting, a specific toy, etc.).
   
2. What types of symbols work best for the dog (arbitrary, pictures, drawing). Taari is very visual - watches TV, sees images on bags. Spice on the other hand isn't and it might be a good test of his eye sight.
   
3. Teach the dog to count. Do I use a number two "2" or represent two with two large dots?
4. Another way to play some of the games that are in Kay Laurence's book Learning Games.
5. Teaching the dogs a sequence. First push A, then B and finally C. This may be a way to exercise the dogs in the house when it is too cold and/or the sidewalks are too dangerous.
   
6. Expanding on the above, as well as a sequence have a "command" attached to the sign. For example when the dog pushes A the sign says "sit", then the dog goes and pushes B the sign says "down", etc. You could almost imagine a rally type course set up with these signs. Of course you may have to program the signs to say "command" pause "release". But I think that would be doable since they can record up to 10 seconds.

Too many choices at this point. I will have to figure out something simple to start with.

Also after thinking about it - do I want the talking signs to talk to me or to the dogs? I think that to explore what the dogs can communicate to me the signs should talk to me. Can I have the signs do both? Perhaps especially if they are "different".