def __init__(self):
print("Loading forest model...")
tic()
self.forest = utilites.loadVariableFromFile(
"static/Corel5K/forest/forest_128.pkl")
print("Done.")
toc()
self.train_vectors = loadmat(
utilites.getAbsPath('static/Corel5K/train_vectors_original.mat'))
self.train_vectors = self.train_vectors['train_vectors']
self.train_file_path = utilites.loadVariableFromFile(
"static/Corel5K/train_file_path.pkl")
# load contents of concepts
self.concepts = utilites.loadVariableFromFile(
"static/Corel5K/cluster_contents.pkl")
self.tag_scores = utilites.loadVariableFromFile(
"static/Corel5K/all_tags_scores.pkl")
self.train_vectors_classic = loadmat(
utilites.getAbsPath(
'static/Corel5K/baseline_features/corel5k_train_feats_classic.mat'
))
self.train_vectors_classic = self.train_vectors_classic[
'corel5k_train_feats_classic']
self.test_vectors_classic = loadmat(
utilites.getAbsPath(
'static/Corel5K/baseline_features/corel5k_test_feats_classic.mat'
))
self.test_vectors_classic = self.test_vectors_classic[
'corel5k_test_feats_classic']
self.test_file_name = utilites.loadVariableFromFile(
'static/Corel5K/corel5k_test_file_name.pkl')
self.feat_dict_classic = dict(
zip(self.test_file_name, self.test_vectors_classic))
# start a matlab session for feature extraction
self.matlab = matlab_wrapper.MatlabSession(
matlab_root="/Applications/MATLAB_R2015b.app") # start matlab
self.matlab.eval('run MatConvNet/matlab/vl_setupnn') # basic config
self.matlab.eval('run vlfeat/toolbox/vl_setup') ## basic config
self.matlab.eval("feature('DefaultCharacterSet', 'UTF8')")
print("Loading cnn model...")
tic()
self.matlab.eval(
"net = load('/Users/TONYSUN/Desktop/SIR_Corel5K_demo/static/cnnmodel/imagenet-matconvnet-vgg-verydeep-16.mat')"
)
toc()
print("Matlab session started.")
print("Ready for work ^_^.")
def getTextVectors():
"""
open the annotation text file and read content
build word vectors using Word2Vec and then extract
the term/vector pairs into a dictionary
:return: the ultimate word vectors
"""
raw_text_file = open(utilites.getAbsPath(setup.corpus_file_path))
raw_text = raw_text_file.readlines()
print("Corpus file " + raw_text_file.name + " was loaded.")
# use re to split the raw text string and replace the original text
# After this all the sentence are split into such format:
# [0]filename, [1]order of annotation, [2]annotation text
raw_text = [
re.split('\t|#', singleLine.replace('\n', ''))
for singleLine in raw_text
]
# now we only need the annotations
annotations = [line[2] for line in raw_text]
# Prepare the sentences
sentences = annotation_to_wordlists(annotations)
# Set values for Word2Vec
num_features = 300 # Use a 300-dimension vector to represent a word
min_word_count = 5 # Word appears less than 5 times will be ignored
num_workers = 4 # Number of threads to run in parallel
context = 5 # Sample 5 words as input for each iteration
# initialize a model using parameters above
word_model = gensim.models.Word2Vec(workers=num_workers,
size=num_features,
min_count=min_word_count,
window=context)
word_model.build_vocab(sentences) # build vocabulary on split sentenced
print("Language model established.")
print("Loading pre-trained language model...")
# initialize the network weights using pre-trained model
word_model.intersect_word2vec_format(utilites.getAbsPath(
setup.lmodel_file_path),
binary=True)
print("Loaded weights from pre-trained Google News language model.")
print("Training models...")
# train the model to get word vectors
word_model.train(sentences)
print("Training completed.")
return extractVecs(word_model)
def generate_forest():
train_original = loadmat(utilites.getAbsPath('Corel5K/train_vectors_original.mat'))
train_original = train_original['train_vectors']
train_label = utilites.loadVariableFromFile(utilites.getAbsPath("Corel5K/train_anno_concept.pkl"))
# prepare data
train_data = Data()
train_data.samples = train_original
train_data.labels = train_label
train_data.orig_sample_indexes = np.array(range(len(train_original)))
train_data.features = np.array(range(np.shape(train_original)[1]))
tic()
rand_forest = generate_random_forest(train_data, 400)
toc()
return rand_forest
def getTextVectors():
"""
open the annotation text file and read content
build word vectors using Word2Vec and then extract
the term/vector pairs into a dictionary
:return: the ultimate word vectors
"""
raw_text_file = open(utilites.getAbsPath(setup.corpus_file_path))
raw_text = raw_text_file.readlines()
print("Corpus file " + raw_text_file.name + " was loaded.")
# use re to split the raw text string and replace the original text
# After this all the sentence are split into such format:
# [0]filename, [1]order of annotation, [2]annotation text
raw_text = [re.split("\t|#", singleLine.replace("\n", "")) for singleLine in raw_text]
# now we only need the annotations
annotations = [line[2] for line in raw_text]
# Prepare the sentences
sentences = annotation_to_wordlists(annotations)
# Set values for Word2Vec
num_features = 300 # Use a 300-dimension vector to represent a word
min_word_count = 5 # Word appears less than 5 times will be ignored
num_workers = 4 # Number of threads to run in parallel
context = 5 # Sample 5 words as input for each iteration
# initialize a model using parameters above
word_model = gensim.models.Word2Vec(
workers=num_workers, size=num_features, min_count=min_word_count, window=context
)
word_model.build_vocab(sentences) # build vocabulary on split sentenced
print("Language model established.")
print("Loading pre-trained language model...")
# initialize the network weights using pre-trained model
word_model.intersect_word2vec_format(utilites.getAbsPath(setup.lmodel_file_path), binary=True)
print("Loaded weights from pre-trained Google News language model.")
print("Training models...")
# train the model to get word vectors
word_model.train(sentences)
print("Training completed.")
return extractVecs(word_model)
def im2res(self, im_path, test_file_name):
"""
read an image file and give the final parsed result
:param im_path: image file path
:return: top5 concept label count and top5 sample count
"""
im_vec = self.extract_feature(utilites.getAbsPath(im_path))
src, srs, im_near_path, rf_bl_time = self.parse_test_sample(im_vec, test_file_name)
return src, srs, im_near_path, rf_bl_time
def im2res(self, im_path, test_file_name):
"""
read an image file and give the final parsed result
:param im_path: image file path
:return: top5 concept label count and top5 sample count
"""
im_vec = self.extract_feature(utilites.getAbsPath(im_path))
src, srs, im_near_path, rf_bl_time = self.parse_test_sample(
im_vec, test_file_name)
return src, srs, im_near_path, rf_bl_time
def __init__(self):
print ("Loading forest model...")
tic()
self.forest = utilites.loadVariableFromFile("static/Corel5K/forest/forest_128.pkl")
print ("Done.")
toc()
self.train_vectors = loadmat(utilites.getAbsPath("static/Corel5K/train_vectors_original.mat"))
self.train_vectors = self.train_vectors["train_vectors"]
self.train_file_path = utilites.loadVariableFromFile("static/Corel5K/train_file_path.pkl")
# load contents of concepts
self.concepts = utilites.loadVariableFromFile("static/Corel5K/cluster_contents.pkl")
self.tag_scores = utilites.loadVariableFromFile("static/Corel5K/all_tags_scores.pkl")
self.train_vectors_classic = loadmat(
utilites.getAbsPath("static/Corel5K/baseline_features/corel5k_train_feats_classic.mat")
)
self.train_vectors_classic = self.train_vectors_classic["corel5k_train_feats_classic"]
self.test_vectors_classic = loadmat(
utilites.getAbsPath("static/Corel5K/baseline_features/corel5k_test_feats_classic.mat")
)
self.test_vectors_classic = self.test_vectors_classic["corel5k_test_feats_classic"]
self.test_file_name = utilites.loadVariableFromFile("static/Corel5K/corel5k_test_file_name.pkl")
self.feat_dict_classic = dict(zip(self.test_file_name, self.test_vectors_classic))
# start a matlab session for feature extraction
self.matlab = matlab_wrapper.MatlabSession(matlab_root="/Applications/MATLAB_R2015b.app") # start matlab
self.matlab.eval("run MatConvNet/matlab/vl_setupnn") # basic config
self.matlab.eval("run vlfeat/toolbox/vl_setup") ## basic config
self.matlab.eval("feature('DefaultCharacterSet', 'UTF8')")
print ("Loading cnn model...")
tic()
self.matlab.eval(
"net = load('/Users/TONYSUN/Desktop/SIR_Corel5K_demo/static/cnnmodel/imagenet-matconvnet-vgg-verydeep-16.mat')"
)
toc()
print ("Matlab session started.")
print ("Ready for work ^_^.")
def show_images(image_list,titles=None):
"""Display a list of images"""
images = []
for j in range(0, len(image_list)):
images.append(io.imread(utilites.getAbsPath(image_list[j])))
n_ims = len(images)
if titles is None: titles = ['(%d)' % i for i in range(1,n_ims + 1)]
fig = plt.figure()
n = 1
for image,title in zip(images,titles):
a = fig.add_subplot(1,n_ims,n) # Make subplot
if image.ndim == 2: # Is image grayscale?
plt.gray() # Only place in this blog you can't replace 'gray' with 'grey'
plt.imshow(image)
a.set_title(title)
n += 1
# fig.set_size_inches(np.array(fig.get_size_inches()) * n_ims)
plt.show()
# return text vectors calculated using Word2Vec by gensim
"""
open the annotation text file and read content
build word vectors using Word2Vec and then extract
the term/vector pairs into a dictionary
"""
# get all filtered term(tag) names
terms_corel5k_filtered = utilites.loadVariableFromFile(
"Corel5k/terms_corel5k_filtered.pkl")
# get training image annotations: lists of separate terms
train_anno_filtered = utilites.loadVariableFromFile(
"Corel5k/train_anno_filtered.pkl")
# initialize a model using parameters above
word_model = gensim.models.Word2Vec.load_word2vec_format(utilites.getAbsPath(
setup.lmodel_file_path),
binary=True)
"""
Calculate similarity matrix using given vectors
We use pairwise distances to build the matrix
"""
print("Extracting word vectors...")
vecs = []
# Index2word is a list that contains the names of the words in
for word in terms_corel5k_filtered:
vecs.append(
word_model[word]) # now we extract all word vectors from the model
print("Term vectors haven been created.")
d_pairwise_vecs = 1 - pairwise.pairwise_distances(vecs, metric='cosine')
print("Similarity matrix has been built.")
from sklearn.metrics import pairwise
# return text vectors calculated using Word2Vec by gensim
"""
open the annotation text file and read content
build word vectors using Word2Vec and then extract
the term/vector pairs into a dictionary
"""
# get all filtered term(tag) names
terms_corel5k_filtered = utilites.loadVariableFromFile("Corel5k/terms_corel5k_filtered.pkl")
# get training image annotations: lists of separate terms
train_anno_filtered = utilites.loadVariableFromFile("Corel5k/train_anno_filtered.pkl")
# initialize a model using parameters above
word_model = gensim.models.Word2Vec.load_word2vec_format(utilites.getAbsPath(setup.lmodel_file_path), binary=True)
"""
Calculate similarity matrix using given vectors
We use pairwise distances to build the matrix
"""
print("Extracting word vectors...")
vecs = []
# Index2word is a list that contains the names of the words in
for word in terms_corel5k_filtered:
vecs.append(word_model[word]) # now we extract all word vectors from the model
print("Term vectors haven been created.")
d_pairwise_vecs = 1 - pairwise.pairwise_distances(vecs, metric="cosine")
print("Similarity matrix has been built.")
# get all terms from txt file
"""
test_file = open(utilites.getAbsPath('static/Corel5K/corel5k_test_list.txt'))
test_file_list = test_file.readlines()
test_file_list = [term.strip().decode('utf-8').replace('\n', '') + '.jpeg' for term in test_file_list]
utilites.saveVariableToFile(test_file_list, utilites.getAbsPath('static/Corel5K/corel5k_test_list.pkl'))
"""
test_file_list = utilites.loadVariableFromFile('static/Corel5K/corel5k_test_list.pkl')
train_file_path = utilites.loadVariableFromFile("static/Corel5K/train_file_path.pkl")
test_anno = utilites.loadVariableFromFile('static/Corel5K/test_anno_filtered.pkl')
train_anno = utilites.loadVariableFromFile('static/Corel5K/train_anno_filtered.pkl')
train_anno_concept = utilites.loadVariableFromFile("static/Corel5K/train_anno_concept.pkl")
test_anno_concept = utilites.loadVariableFromFile("static/Corel5K/test_anno_concept.pkl")
all_prob = utilites.loadVariableFromFile("static/Corel5K/all_probs.pkl")
concepts = utilites.loadVariableFromFile("static/Corel5K/cluster_contents.pkl")
train_vectors = loadmat(utilites.getAbsPath('static/Corel5K/train_vectors_original.mat'))
train_vectors = train_vectors['train_vectors']
test_vectors = loadmat(utilites.getAbsPath('static/Corel5K/test_vectors_original.mat'))
test_vectors = test_vectors['test_vectors']
train_vectors_classic = loadmat(utilites.getAbsPath('static/Corel5K/baseline_features/corel5k_train_feats_classic.mat'))
train_vectors_classic = train_vectors_classic['corel5k_train_feats_classic']
test_vectors_classic = loadmat(utilites.getAbsPath('static/Corel5K/baseline_features/corel5k_test_feats_classic.mat'))
test_vectors_classic = test_vectors_classic['corel5k_test_feats_classic']
print("Loading forest model...")
tic()
forest = utilites.loadVariableFromFile("static/Corel5K/forest/forest_128.pkl")
a_rs = []
for tree in forest:
rc, rs = parse_single_tree(sample, tree)
a_rc.append(rc.label_count)
a_rs = a_rs + list(rs)
a_rc = np.asarray(a_rc)
sum_a_rc = np.sum(a_rc, axis=0) # get count in all trees for each concept
sum_a_rs = Counter(a_rs)
return sum_a_rc, sum_a_rs
# test code here
test_original = loadmat(
utilites.getAbsPath('Corel5K/test_vectors_original.mat'))
test_original = test_original['test_vectors']
test_sample = test_original[0]
forest = load_forest()
# rc, rs = parse_single_tree(test_sample, forest[0])
# src: sum of concept count, srs: sum of retrieved sample count
src, srs = parse_forest(test_sample, forest)
label_name = range(100)
src_dict = dict(zip(label_name, src))
src_sorted = sorted(src_dict.items(), key=itemgetter(1))[::-1]
srs_sorted = sorted(srs.items(), key=itemgetter(1))[::-1]
src_top10 = src[0:10]
srs_top10 = srs_sorted[0:10]
def toc(tempBool=True):
# Prints the time difference yielded by generator instance TicToc
tempTimeInterval = next(TicToc)
if tempBool:
print( "Elapsed time: %f seconds.\n" %tempTimeInterval )
def tic():
# Records a time in TicToc, marks the beginning of a time interval
toc(False)
"""
Prepare data
"""
# START TO RUN HERE
# load datasets
train_original = loadmat(utilites.getAbsPath('Corel5K/train_vectors_original.mat'))
test_original = loadmat(utilites.getAbsPath('Corel5K/test_vectors_original.mat'))
train_original = train_original['train_vectors']
test_original = test_original['test_vectors']
# l2 normalize the feature vectors
train_original_list = [tr.astype('float32') for tr in train_original]
# construct a pairwise distance matrix to lookup (affinity matrix)
d_cos_train_vecs = 1 - pairwise.pairwise_distances(train_original_list, metric='cosine')
# d_eu_train_vecs = pairwise.pairwise_distances(train_pca_300_list, metric='euclidean')
# get all terms from txt file
terms_file = open(utilites.getAbsPath('Corel5K/corel5k_words.txt'))
terms_corel5k = terms_file.readlines()
print("Term file " + terms_file.name + " was loaded.")
if tempBool:
print("Elapsed time: %f seconds.\n" % tempTimeInterval)
def tic():
# Records a time in TicToc, marks the beginning of a time interval
toc(False)
"""
Prepare data
"""
# START TO RUN HERE
# load datasets
train_original = loadmat(
utilites.getAbsPath('Corel5K/train_vectors_original.mat'))
test_original = loadmat(
utilites.getAbsPath('Corel5K/test_vectors_original.mat'))
train_original = train_original['train_vectors']
test_original = test_original['test_vectors']
# l2 normalize the feature vectors
train_original_list = [tr.astype('float32') for tr in train_original]
# construct a pairwise distance matrix to lookup (affinity matrix)
d_cos_train_vecs = 1 - pairwise.pairwise_distances(train_original_list,
metric='cosine')
# d_eu_train_vecs = pairwise.pairwise_distances(train_pca_300_list, metric='euclidean')
# get all terms from txt file
terms_file = open(utilites.getAbsPath('Corel5K/corel5k_words.txt'))
a_rc = []
a_rs = []
for tree in forest:
rc, rs = parse_single_tree(sample, tree)
a_rc.append(rc.label_count)
a_rs = a_rs + list(rs)
a_rc = np.asarray(a_rc)
sum_a_rc = np.sum(a_rc, axis=0) # get count in all trees for each concept
sum_a_rs = Counter(a_rs)
return sum_a_rc, sum_a_rs
# test code here
test_original = loadmat(utilites.getAbsPath('Corel5K/test_vectors_original.mat'))
test_original = test_original['test_vectors']
test_sample = test_original[0]
forest = load_forest()
# rc, rs = parse_single_tree(test_sample, forest[0])
# src: sum of concept count, srs: sum of retrieved sample count
src, srs = parse_forest(test_sample, forest)
label_name = range(100)
src_dict = dict(zip(label_name, src))
src_sorted = sorted(src_dict.items(), key=itemgetter(1))[::-1]
srs_sorted = sorted(srs.items(), key=itemgetter(1))[::-1]
src_top10 = src[0:10]
srs_top10 = srs_sorted[0:10]
