tony607 / keras_age_gender Goto Github PK

The output varies too quickly to be interpreted well at times. An easy solution would be to take a 100 frames and print the average value of the predictions, but I was wondering how this would do against taking an average embedding of, say 100 frames, and run the prediction on that average image.

I try to get the embeddings of the images by storing the outputs of the final dense layer, however this slows down the system considerably. I notice that results = self.model.predict(face_imgs) however doesn't do so.

Is there a better way to achieve this?

thank you for your share

Can you tell me the accuracy of the model?

Can you please share me your suggestions to train custom age and gender detection.

The reason to ask is, when I tested the pre trained model, I get wrong detections, Female,21 is recognised as Male,21.
Thanks
Guru

When I run
python realtime_demo.py

I get an error:
cv2.error: OpenCV(3.4.1) /io/opencv/modules/objdetect/src/cascadedetect.cpp:1698: error: (-215) !empty() in function detectMultiScale

Hi,

I have a problem with the rectangle and labels. It is not drawing continuously as your post, do you know why is that??

Can you please provide train.py file and Data set for the above problem.???

i try test this
but on install requirement
give this error
Could not find a version that satisfies the requirement opencv==1.0.1 (from -r requirements.txt (line 3)) (from versions: )
No matching distribution found for opencv==1.0.1 (from -r requirements.txt (line 3))
on run demo return this:
Using TensorFlow backend.
Illegal instruction

When I run the realtime_demo.py,there is not the lable and the rectangular

Hello everyone.I got this error in my raspberry pi.How to i fix it?

Using TensorFlow backend. Traceback (most recent call last): File "realtime_demo.py", line 143, in <module> main() File "realtime_demo.py", line 138, in main face = FaceCV(depth=depth, width=width) File "realtime_demo.py", line 27, in __init__ self.model = WideResNet(face_size, depth=depth, k=width)() File "/home/pi/Desktop/Keras_age_gender-master/wide_resnet.py", line 122, in __call__ use_bias=self._use_bias)(inputs) # "One conv at the beginning (spatial size: 32x32)" File "/home/pi/.local/lib/python3.5/site-packages/keras/engine/topology.py", line 619, in __call__ output = self.call(inputs, **kwargs) File "/home/pi/.local/lib/python3.5/site-packages/keras/layers/convolutional.py", line 168, in call dilation_rate=self.dilation_rate) File "/home/pi/.local/lib/python3.5/site-packages/keras/backend/tensorflow_backend.py", line 3329, in conv2d x = tf.nn.convolution( AttributeError: module 'tensorflow.python.ops.nn' has no attribute 'convolution'

I see that you are downloading weights.hdf5 file from https://github.com/Tony607/Keras_age_gender/releases/download/V1.0/release

What exactly does this file have ?

Is this like a pre trained model to predict age or have you done any external training to train any model ?

Please let me know so that I can understand the code