Having recently delved into the realms of keras and machine learning, I am experiencing some issues with my code that I believe stem from my model. My objective is to train a model using a CSV file where each row represents a flattened image, essentially making each row of data an individual image. However, I keep encountering the following error message:
ValueError: Exception encountered when calling layer "sequential" (type Sequential).
Input 0 of layer "dense3" is incompatible with the layer: expected min_ndim=2, found ndim=1. Full shape received: (4096,)
Call arguments received by layer "sequential" (type Sequential):
• inputs=tf.Tensor(shape=(4096,), dtype=uint8)
• training=False
• mask=None
I have explored various online resources and attempted multiple solutions, but being new to this field, there are still aspects that elude me. The dimensions of the image are 64x64, leading me to assume that the columns represent the input size. As a result, I set the input size to 64*64 and adjusted the predict batch size accordingly to ensure compatibility with the flatten image for real-time prediction. This approach resolved the issue for the initial layer (input layer), however, subsequent dense layers are proving problematic. Additionally, I have positioned the image label at the beginning of each row in the CSV file.
def modelTraining():
model = keras.models.Sequential()
model.add(keras.Input(shape=(4096,)))
model.add(keras.layers.Dense(4096, activation='relu', name = "dense2"), )
model.add(keras.layers.Dense(1, activation='softmax', name = "dense3", ))
model.compile(optimizer='rmsprop',
loss='binary_crossentropy',
metrics=['accuracy'])
#Retrieve training image data
trainingImageMatrix = []
trainingLabelsArray = []
temp = 0
with open(r'C:\Users\Jason\Desktop\signLanguage\testOpenCV\training.csv', 'r', newline='') as myFile:
reader = csv.reader(myFile, delimiter=',')
for row in reader:
temp+=1
imageDataToAdd = np.zeros(4096)
for i in range (1, 4097):
imageDataToAdd[i - 1] = row[i]
trainingLabelsArray.append(int(row[0])) #Extracting the first index of each row
trainingImageMatrix.append(imageDataToAdd)
print("reading row: "+str(temp))
#if temp >=10:
#break
trainingImageMatrix = np.array(trainingImageMatrix)
trainingLabelsArray = np.array(trainingLabelsArray)
model.fit(trainingImageMatrix, trainingLabelsArray, epochs = 20, batch_size = 4096)
return model
model.predict(grayFrame.flatten(), batch_size = 4096)
The above line appears in another segment of my code, where it currently serves as my prediction method.
I would greatly appreciate guidance on how to accurately specify the size and tailor the layers according to the requirements of my project.