Styjun

The code below is from coursera

import numpy as np
from google.colab import files
from keras.preprocessing import image

uploaded = files.upload()

for fn in uploaded.keys():

 # predicting images
 path = '/content/' + fn
 img = image.load_img(path, target_size=(300, 300))
 x = image.img_to_array(img)
 x = np.expand_dims(x, axis=0)

 images = np.vstack([x])
 classes = model.predict(images, batch_size=10)
 print(classes[0])
 if classes[0]>0.5:
 print(fn + " is a human")
 else:
 print(fn + " is a horse")

The code below is from stackoverflow

# step 1
filenames = tf.constant(['im_01.jpg', 'im_02.jpg', 'im_03.jpg', 'im_04.jpg'])
labels = tf.constant([0, 1, 0, 1])

# step 2: create a dataset returning slices of `filenames`
dataset = tf.data.Dataset.from_tensor_slices((filenames, labels))

# step 3: parse every image in the dataset using `map`
def _parse_function(filename, label):
 image_string = tf.read_file(filename)
 image_decoded = tf.image.decode_jpeg(image_string, channels=3)
 image = tf.cast(image_decoded, tf.float32)
 return image, label

dataset = dataset.map(_parse_function)
dataset = dataset.batch(2)

# step 4: create iterator and final input tensor
iterator = dataset.make_one_shot_iterator()
images, labels = iterator.get_next()

The code below is the handwritten digits for my model using a cnn.

import tensorflow as tf
from tensorflow import keras
import numpy as np


mnist = tf.keras.datasets.mnist
(training_images, training_labels), (test_images, test_labels) = mnist.load_data()

training_images=training_images.reshape(60000, 28, 28, 1)
training_images=training_images / 255.0
test_images = test_images.reshape(10000, 28, 28, 1)
test_images=test_images/255.0

class myCallback(tf.keras.callbacks.Callback):
 def on_epoch_end(self, epoch, logs=):
 if(logs.get('acc')>=0.998):
 print("nReached 99.8% accuracy so cancelling training!")
 self.model.stop_training = True

callbacks = myCallback()

model = tf.keras.models.Sequential([
 tf.keras.layers.Conv2D(64, (3,3), activation='relu', input_shape=(28, 28, 1)),
 tf.keras.layers.MaxPooling2D(2, 2),
 tf.keras.layers.Flatten(),
 tf.keras.layers.Dense(128, activation='relu'),
 tf.keras.layers.Dense(10, activation='softmax')
])

model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model.fit(training_images, training_labels, epochs=20, callbacks=[callbacks])

SO, I have seen this post regarding training on my own data (images), however, this is not for the current version of tensorflow==2.0.0.0alpha0 in which I am trying to write code in.

The purpose of the program is to classify a single image that I upload (such as a drawing/writing of my own handwritten digits and see if my model & my handwriting is legible to the computer. I will not being using multiple number values, nor letters as in alpha-numeric, just simply digits (singular integers) for now. I have already created a cnn that achieves the state of the art (99.8%) correct classification on the model however, I am still questioning how to do run this against files that I upload, or from the disk. The code that I have provided from coursera gives an example of how to do this for using an upload image button, however, this is running files from google drive, and I would like to do so on from my own disk as I am not using colab/google to run this project.
Thank you for your time, as I hope the answer helps others as well.
Cody Quist