Skip to content

Instantly share code, notes, and snippets.

Embed
What would you like to do?
Lenet implementation for Acceleration in Keras
# The model is inspired from the LeNet, 1998 paper by Le Cunn
# Credits: https://github.com/keras-team/keras/blob/master/examples/mnist_cnn.py
from __future__ import print_function
import keras
from keras.datasets import mnist
from keras.models import Sequential
from keras.layers import Dense, Dropout, Flatten
from keras.layers import Conv2D, MaxPooling2D
from keras import backend as K
from keras.layers.normalization import BatchNormalization
batch_size = 128
epochs = 30
model = Sequential()
model.add(Conv2D(256, kernel_size=(3, 3),
activation='relu',
input_shape=input_shape)) #Convolution
model.add(MaxPooling2D(pool_size=(2, 2))) #Subsampling
model.add(Dropout(0.25))
model.add(Conv2D(128, (3, 3), activation='relu')) #Convolution
model.add(MaxPooling2D(pool_size=(2, 2))) #Subsampling
model.add(Dropout(0.25))
model.add(Flatten())
model.add(Dense(128, activation='relu')) # Full Connection
model.add(Dropout(0.5))
model.add(Dense(64, activation='relu')) # Full Connection
model.add(Dropout(0.5))
model.add(Dense(1))
model.compile(loss='mse',#loss=keras.losses.categorical_crossentropy,
optimizer=keras.optimizers.Adadelta(),
metrics=['mean_squared_error', 'mean_absolute_error', 'mean_absolute_percentage_error', 'cosine_proximity'])
# train_xs, train_accels, val_xs, val_accels
history=model.fit(train_xs, train_accels,
batch_size=batch_size,
epochs=epochs,
verbose=1,
validation_data=(val_xs, val_accels))
score = model.evaluate(val_xs, val_accels, verbose=0)
# To try predict acceleration using train data
pred = model.predict(train_xs)
#Print acceleration first n frames
for i in pred[0:500]:
print(i*180/scipy.pi)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment