-
Notifications
You must be signed in to change notification settings - Fork 32
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Using the main function to train and test. #3
Comments
Do you now have more datasets available? |
Can you show an example of how to use recognize_gesture()? |
i have trained that asl mnist dataset from kaggle and i got 95 percent accuracy.but when i trid with your image i dint get proper output canu please help me to get out of this. |
Hi Karthick, could you share the link to the asl mnist dataset? |
I'm getting |
Could you try to print the loading data and check whether that is in
correct format?
…On Thu, Apr 2, 2020, 5:54 PM TusharAI ***@***.***> wrote:
I'm getting
*le = loadClassifier('label_en (1).pkl')* error in pipeline_final.ipynb
file
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#3 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AGFNX43HNUKSZKQT7OE2NGDRKR7ZPANCNFSM4DXADTOA>
.
|
Firstly apologies for having put out some code that is not ready to use (i.e., a one click solution). I have received a lot of queries on how to run the main method, I've decided to share the same here.
The code here was never meant to be a one-click solution to static gesture recognition and in fact, was shared so that the methods used to train the models were available. Either way, if you are interested in running the same as opposed to using the methods in an independent program, you can try the following:
Ignore lines 505 to 508. Change 509, 512 and 513 appropriately to include the correct list(s) of users (as per the data you have downloaded).
Add a line:
gs = GestureRecognizer('/path/to/dataset/')
(provide the correct path here)This constructor is different from the one we've used. The reason we used the other one was because we had already trained the models. Using this constructor means that the models will get trained when you call the train method.
Then use the following as it is:
gs.train(user_tr) gs.save_model(name = "your-model-name.pkl.gz", version = "0.0.1", author = 'ss') print "The GestureRecognizer is saved to disk"
Once it is trained (it might take a lot of time to train), your model will be saved to disk. Then on, you can simply load the model and use it to test (i.e., detect gestures) using the recognize_gesture() method.
For this,
gs = GestureRecognizer.load_model(name = "your-model-name.pkl.gz") # automatic dict unpacking gs.recognize_gesture()
Hope this helps!
The text was updated successfully, but these errors were encountered: