This App is No Longer Maintained. Visit the Replacement at StrokesPlus.net

StrokesPlus Forum
                       
StrokesPlus Forum
Home | Profile | Register | Active Topics
Members | Search | FAQ
 All Forums
 Feature Updates
 Closed or Deferred Requests
 [CLOSED] more robust gesture recognition

Note: You must be registered in order to post a reply.
To register, click here. Registration is FREE!

UserName:
Password:
Format Mode:
Format: BoldItalicizedUnderlineStrikethrough Align LeftCenteredAlign Right Horizontal Rule Insert HyperlinkInsert EmailInsert Image Insert CodeInsert QuoteInsert List
   
Message:

* HTML is OFF
* Forum Code is ON
Smilies
Smile [:)] Big Smile [:D] Cool [8D] Blush [:I]
Tongue [:P] Evil [):] Wink [;)] Clown [:o)]
Black Eye [B)] Eight Ball [8] Frown [:(] Shy [8)]
Shocked [:0] Angry [:(!] Dead [xx(] Sleepy [|)]
Kisses [:X] Approve [^] Disapprove [V] Question [?]

 
Check here to subscribe to this topic.
   

T O P I C    R E V I E W
tr4np Posted - 04/29/2015 : 13:59:41
S+ is great, but I do sometimes see issues where my gesture is misinterpreted. I'm sure it's my fault for not giving a better drawing during the training, but perhaps the training could be improved. Might I suggest having the user draw the gesture multiple times, and somehow combine the data from the multiple training gestures into an "average" that is more likely to be consistently recognized? Alternatively, you could allow multiple gestures for a given Action.

For example, my "down" gesture is never straight down. It's always wavy (I often use a Wacom tablet instead of a mouse) and sometimes points toward one side, causing it to be recognized as a bottom-left or bottom-right gesture instead. If I were able to use three different gestures for my "down" action (e.g. straight down, slightly left, and slightly right), S+ could more consistently identify my intended action.
3   L A T E S T    R E P L I E S    (Newest First)
Rob Posted - 04/29/2015 : 15:08:44
You're welcome! I'm glad you're enjoying it :-)
tr4np Posted - 04/29/2015 : 15:05:24
Ah, I did not realize that's how training mode worked. I simply created my new [single] gestures from the Actions window. My apologies for failing to RTFM!

I must say, your support has been awesome! Thank you for providing an excellent product (for free!).
Rob Posted - 04/29/2015 : 14:54:57
You can enter Training Mode (see Help) and train S+ to recognize better based on your drawings. When a gesture is drawn, S+ will go through each gesture pattern for all gestures (Training Mode saves as many as you train) ranking on the highest match probability, then fires that action.

So, enter Training mode, draw a down and slightly left and let go, then S+ will pop up a box with what it guessed. If it's not what you wanted, enter the name of the gesture you intended and it will add that pattern to the gesture name entered. In this case, let's say it recognized it as "Down-Left" and you cleared that out and typed "Down". S+ will add that pattern to the "Down" gesture points (now "Down" has 2 patterns associated with it). Repeat it again for down and slightly right, entering "Down", now S+ has 3 patterns for the "Down" gesture.

The only thing I would like to mention is that each learned pattern adds to the overall processing time for each time S+ is trying to match (after you let go of the stroke button), so I don't recommend doing this 100 times as you may start to notice the lag as S+ tests the gesture against 100 "Down" patterns. However, just so you know, if you delete a gesture and recreate it, you start back at 1 pattern for that gesture, so you can always reset.

StrokesPlus Forum © 2011-2018 Rob Yapchanyk Go To Top Of Page
Snitz Forums 2000