gabrieltseng

gabrieltseng

gabrieltseng

import torch
import torch.nn as nn
from torch.autograd import Variable, Function
import random

class WARP(Function): 
    '''
    autograd function of WARP loss
    '''
    @staticmethod

gabrieltseng

# Define a function to train SVM
def train_svm(data, labels, kernel, C, gamma, degree, coef0):
    """A generic SVM training function, with arguments based on the chosen kernel."""
    if kernel == 'linear':
        model = SVC(kernel=kernel, C=C)
    elif kernel == 'poly':
        model = SVC(kernel=kernel, C=C, degree=degree, coef0=coef0)
    elif kernel == 'rbf':
        model = SVC(kernel=kernel, C=C, gamma=gamma)
    else:

gabrieltseng

from keras.applications.inception_v3 import InceptionV3
from keras.models import Model
from keras.layers import Dense, GlobalAveragePooling2D

base_model = InceptionV3(weights='imagenet', include_top=False)

# Adding a 'top' which outputs 2 categories

x = base_model.output
x = GlobalAveragePooling2D()(x)

gabrieltseng

def make_xy(model, path):
    
    #making the X values
    for z in range(len(model)):
    
        gen = image.ImageDataGenerator()

        batches = gen.flow_from_directory(path, 
                                          target_size = (224, 224), 
                                          batch_size = 32, 

gabrieltseng

def ResNet_inv():
    RN_mean = np.array([123.68, 116.779, 103.939], dtype=np.float32).reshape((3,1,1))
    
    def RN_preprocess(x, RN_mean):
        x = x - np.asarray(RN_mean)
        return x[:, ::-1] # reverse axis bgr->rgb
    
    img_input = Input(shape =(3,224,224))
    bn_axis = 1
    

gabrieltseng

def vgg_bn(dropout = 0.5, BN = True):
    #returns a vgg model with batch normalization and a defined dropout
    #the weights are adjusted to whichever dropout you select
    prev_vgg = load_model('/home/ubuntu/invasive_vgg.model')
    
    prev_layers = prev_vgg.layers[:-5]
    new_vgg = Sequential(prev_layers)
    
    def new_FCL(model, num_filters, dropout):
        # In Keras, a fully connected layer is called 'Dense' 

gabrieltseng

from keras.applications.resnet50 import ResNet50

ResNet = ResNet50(weights='imagenet')