如何在 Python 程式中使用 Google Teachable Machine 建立的 model
# Python program to control mouse based on head position # Import necessary modules import numpy as np import cv2 import tensorflow as tf # Using laptop’s webcam as the source of video cap = cv2.VideoCapture(0) cap.set(3, 480) # ID 3 = width cap.set(4, 320) # ID 4 = height # Labels — The various outcome possibilities labels = ["Up","Down","Left","Right","Neutral"] # Loading the model weigths we just downloaded model = tf.keras.models.load_model("keras_model.h5", compile = False) while True: success, image = cap.read() if success == False: break # Necessary to avoid conflict between left and right image = cv2.flip(image, 1) cv2.imshow("Frame", image) # The model takes an image of dimensions (224,224) as input so let’s # reshape our image to the same. img = cv2.resize(image, (224, 224)) # Convert the image to a numpy array img = np.array(img, dtype=np.float32) img = np.expand_dims(img, axis=0) # Normalizing input image img = img / 255 # Predict the class prediction = model.predict(img) # Map the prediction to the labels # Rnp.argmax returns the indices of the maximum values along an axis. predicted_labels = labels[np.argmax(prediction[0], axis=-1)] # print(predicted_labels) print(predicted_labels, np.argmax(prediction[0], axis=-1), prediction[0]) # Close all windows if one second has passed and ‘q’ is pressed if cv2.waitKey(1) & 0xFF == ord('q'): break # Release open connections cap.release() cv2.destroyAllWindows()
# Python program to control mouse based on head position # Import necessary modules import numpy as np import cv2 import tensorflow as tf # Using laptop’s webcam as the source of video cap = cv2.VideoCapture(0) cap.set(3, 480) # ID 3 = width cap.set(4, 320) # ID 4 = height # Labels — The various outcome possibilities labels = ["Up","Down","Left","Right","Neutral"] # Loading the model weigths we just downloaded model = tf.keras.models.load_model("keras_model.h5", compile = False) while True: success, image = cap.read() if success == False: break # Necessary to avoid conflict between left and right image = cv2.flip(image, 1) cv2.imshow("Frame", image) # The model takes an image of dimensions (224,224) as input so let’s # reshape our image to the same. img = cv2.resize(image, (224, 224)) # Convert the image to a numpy array img = np.array(img, dtype=np.float32) img = np.expand_dims(img, axis=0) # Normalizing input image img = img / 255 # Predict the class prediction = model.predict(img) # Map the prediction to the labels # Rnp.argmax returns the indices of the maximum values along an axis. predicted_labels = labels[np.argmax(prediction[0], axis=-1)] # print(predicted_labels) print(predicted_labels, np.argmax(prediction[0], axis=-1), prediction[0]) # Close all windows if one second has passed and ‘q’ is pressed if cv2.waitKey(1) & 0xFF == ord('q'): break # Release open connections cap.release() cv2.destroyAllWindows()