Closed willy90806k closed 6 years ago
You have 8 sensors, and each sensor return 8 values.
Can i ask you how to take 64 value by your method? I'm a myo's novice. I think mine just a mistake.
@willy90806k just make one dimension array from your output data. And don't forget to remove date value.
I just only can take 8 value out in my EMG extractor. Is it using the EMG data only ? Or also have another.
If you have some examples to me , i will be grateful.
@willy90806k post output
@exelban I post my extractor at Google Driver ,you can try it. It will put your output values at [\8channelChart\8channelChart \ .txt] .
https://drive.google.com/file/d/1PQp0uiMvHocPObLMQ2ameY7_DtZigYvI/view?usp=sharing
0 6 -4 -2 7 29 34 -1
0 -15 -3 -5 4 -49 -22 0
1 2 -4 -1 -12 40 -6 -1
-1 3 1 0 -5 16 19 -1
-2 -3 -1 -3 0 -85 -18 1
-2 -7 -1 3 4 102 2 -1
It look like this , and I change it to the csv in tensorflow then change it to npx .
@willy90806k ok, sorry. I just get 8 times value from sensor for some unification. So, 64 array its just collected 8 times EMG data from myo-armband.
I have armband only for few days for tests, and forgot about this. I will try to get armband for few days and post code for collecting data.
Please don't say sorry . I'm grateful you can answer . :)) [ def model(_NUM_CLASSES = 4):
with tf.name_scope('data'):
x = tf.placeholder(tf.float32, shape=[None, 64], name='Input')
y = tf.placeholder(tf.float32, shape=[None, _NUM_CLASSES], name='Output')
# Store layers weight & bias
weights = {
'h1': tf.Variable(tf.random_normal([64, 528])),
'h2': tf.Variable(tf.random_normal([528, 786])),
'h3': tf.Variable(tf.random_normal([786, 1248])),
'out': tf.Variable(tf.random_normal([1248, _NUM_CLASSES]))
] So , if I change 64 to 8 ,should i change weight's number or _NUM_CLASSES ?
@willy90806k nope. You need to change all numbers of neurones.
Why you can have 64 features to use? Myo just have 8 feature , doesn't it?