Commit 304bbb57 authored by Aleksis's avatar Aleksis
Browse files

Init commit

parents
Pipeline #64 failed with stages
coverart
db
db-raw
models
output
test
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "{}"
replaced with your own identifying information. (Don't include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright {yyyy} {name of copyright owner}
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
# Similar images finder (using Autoencoders)
Given a set of query images and a set of store inventory images, we find the top-k similar inventory images that are the most 'similar' to the set of query images in an unsupervised way of training an autoencoder, then using its encoder to embed the images and perform kNN in to find 'similar' images.
<p align="center">
<img src="https://github.com/ankonzoid/artificio/blob/master/similar_images_AE/coverart/coverart.jpg" width="60%">
</p>
In this code, we train a convolutional autoencoder on 36 steakhouse food images (6 of each of steak, potato, french fries, salads, burger, asparagus), and make similar image food recommendations based on the above algorithm to achieve a result of:
<p align="center">
<img src="https://github.com/ankonzoid/artificio/blob/master/similar_images_AE/output/result_burger_test.png" width="50%">
</p>
<p align="center">
<img src="https://github.com/ankonzoid/artificio/blob/master/similar_images_AE/output/result_salad_test.png" width="50%">
</p>
The model performs fairly well as a vanilla model with minimal fine-tuned training, in the sense that the top similar recommended images tend to be in same food category as the query image (i.e. querying a burger gives mostly burgers, and querying a salad gives mostly salads, ...). There is still much room for improvement in terms different neural network architectures, more/different training images, hyperparameter tuning to improve the generality of this model.
The algorithm:
1) Train an autoencoder with training images in the same domain as the inventory images
2) Use the trained encoder to embed both the query images and the inventory images
3) Perform kNN (euclidean/cosine similarity) to find the inventory nearest neighbour image embeddings to the query image embeddings, and keep the k closest embeddings as the top-k recommendations
### Usage:
To make sure our similar images finder (trained on steakhouse food images) works on our test images,
1. Run the command:
> python similar_images_AE.py
When the run is complete, your answer images can be found in the `output` directory. However, if you would like to train the model from scratch then:
1. In `similar_images_AE.py`, set:
* `model_name` to either `"simpleAE"` (1 FC hidden layer) or `"convAE"` (CNN)
* `process_and_save_images = True` to perform the proper pre-processing of the images
* `train_model = True` to instruct the program to train the model from scratch (it also saves it once the training is complete)
2. Run the command:
> python similar_images_AE.py
### Required libraries:
* numpy, matplotlib, pylab, sklearn, keras, h5py, pillow
### Authors:
Anson Wong
\ No newline at end of file
'''
similar_images_AE.py (author: Anson Wong / git: ankonzoid)
Image similarity recommender system using an autoencoder-clustering model.
Autoencoder method:
1) Train an autoencoder (simple/Conv) on training images in 'db/images_training'
2) Saves trained autoencoder, encoder, and decoder to 'db/models'
Clustering method:
3) Using our trained encoder in 'db/models', we encode inventory images in 'db/images_inventory'
4) Train kNN model using encoded inventory images
5) Encode query images in 'query', and predict their NN using our trained kNN model
6) Compute a score for each inventory encoding relative to our query encoding (centroid/closest)
7) Make k-recommendations by cloning top-k inventory images into 'answer'
'''
import sys, os, shutil
import numpy as np
sys.path.append("src")
from autoencoders.AE import AE
from clustering.KNN import KNearestNeighbours
from utilities.image_utilities import ImageUtils
from utilities.sorting import find_topk_unique
from utilities.plot_utilities import PlotUtils
import matplotlib.pyplot as plt
def main():
# ========================================
# Set run settings
# ========================================
# Choose autoencoder model
#model_name = "simpleAE"
model_name = "convAE"
process_and_save_images = False # image preproc: resize images and save?
train_autoencoder = False # train from scratch?
# ========================================
# Automated pre-processing
# ========================================
## Set flatten properties ###
if model_name == "simpleAE":
flatten_before_encode = True
flatten_after_encode = False
elif model_name == "convAE":
flatten_before_encode = False
flatten_after_encode = True
else:
raise Exception("Invalid model name which is not simpleAE/convAE")
img_shape = (100, 100) # force resize -> (ypixels, xpixels)
ratio_train_test = 0.8
seed = 100
loss = "binary_crossentropy"
optimizer = "adam"
n_epochs = 100
batch_size = 128
save_reconstruction_on_load_model = False
### KNN training parameters ###
n_neighbors = 5 # number of nearest neighbours
metric = "cosine" # kNN metric (cosine only compatible with brute force)
algorithm = "brute" # search algorithm
recommendation_method = 2 # 1 = centroid kNN, 2 = all points kNN
output_mode = 1 # 1 = output plot, 2 = output inventory db image clones
# ========================================
# Generate expected file/folder paths and settings
# ========================================
# Assume project root directory to be directory of file
project_root = os.path.dirname(__file__)
print("Project root: {0}".format(project_root))
# Query and answer folder
query_dir = os.path.join(project_root, 'test')
answer_dir = os.path.join(project_root, 'output')
# In database folder
db_dir = os.path.join(project_root, 'db')
img_train_raw_dir = os.path.join(project_root, 'db-raw')
img_inventory_raw_dir = os.path.join(project_root, 'db-raw')
img_train_dir = os.path.join(db_dir)
img_inventory_dir = os.path.join(db_dir)
# Run output
models_dir = os.path.join('models')
# Set info file
info = {
# Run settings
"img_shape": img_shape,
"flatten_before_encode": flatten_before_encode,
"flatten_after_encode": flatten_after_encode,
# Directories
"query_dir": query_dir,
"answer_dir": answer_dir,
"img_train_raw_dir": img_train_raw_dir,
"img_inventory_raw_dir": img_inventory_raw_dir,
"img_train_dir": img_train_dir,
"img_inventory_dir": img_inventory_dir,
# Run output
"models_dir": models_dir
}
# Initialize image utilities (and register encoder)
IU = ImageUtils()
IU.configure(info)
# Initialize plot utilities
PU = PlotUtils()
# ========================================
#
# Pre-process save/load training and inventory images
#
# ========================================
# Process and save
if process_and_save_images:
# Training images
IU.raw2resized_load_save(raw_dir=img_train_raw_dir,
processed_dir=img_train_dir,
img_shape=img_shape)
# Inventory images
IU.raw2resized_load_save(raw_dir=img_inventory_raw_dir,
processed_dir=img_inventory_dir,
img_shape=img_shape)
# ========================================
#
# Train autoencoder
#
# ========================================
# Set up autoencoder base class
MODEL = AE()
MODEL.configure(model_name=model_name)
if train_autoencoder:
print("Training the autoencoder...")
# Generate naming conventions
dictfn = MODEL.generate_naming_conventions(model_name, models_dir)
MODEL.start_report(dictfn) # start report
# Load training images to memory (resizes when necessary)
x_data_all, all_filenames = \
IU.raw2resizednorm_load(raw_dir=img_train_dir, img_shape=img_shape)
print("\nAll data:")
print(" x_data_all.shape = {0}\n".format(x_data_all.shape))
# Split images to training and validation set
x_data_train, x_data_test, index_train, index_test = \
IU.split_train_test(x_data_all, ratio_train_test, seed)
print("\nSplit data:")
print("x_data_train.shape = {0}".format(x_data_train.shape))
print("x_data_test.shape = {0}\n".format(x_data_test.shape))
# Flatten data if necessary
if flatten_before_encode:
x_data_train = IU.flatten_img_data(x_data_train)
x_data_test = IU.flatten_img_data(x_data_test)
print("\nFlattened data:")
print("x_data_train.shape = {0}".format(x_data_train.shape))
print("x_data_test.shape = {0}\n".format(x_data_test.shape))
# Set up architecture and compile model
MODEL.set_arch(input_shape=x_data_train.shape[1:],
output_shape=x_data_train.shape[1:])
MODEL.compile(loss=loss, optimizer=optimizer)
MODEL.append_arch_report(dictfn) # append to report
# Train model
MODEL.append_message_report(dictfn, "Start training") # append to report
MODEL.train(x_data_train, x_data_test,
n_epochs=n_epochs, batch_size=batch_size)
MODEL.append_message_report(dictfn, "End training") # append to report
# Save model to file
MODEL.save_model(dictfn)
# Save reconstructions to file
MODEL.plot_save_reconstruction(x_data_test, img_shape, dictfn, n_plot=10)
else:
# Generate naming conventions
dictfn = MODEL.generate_naming_conventions(model_name, models_dir)
# Load models
MODEL.load_model(dictfn)
# Compile model
MODEL.compile(loss=loss, optimizer=optimizer)
# Save reconstructions to file
if save_reconstruction_on_load_model:
x_data_all, all_filenames = \
IU.raw2resizednorm_load(raw_dir=img_train_dir, img_shape=img_shape)
if flatten_before_encode:
x_data_all = IU.flatten_img_data(x_data_all)
MODEL.plot_save_reconstruction(x_data_all, img_shape, dictfn, n_plot=10)
# ========================================
#
# Perform clustering recommendation
#
# ========================================
# Load inventory images to memory (resizes when necessary)
x_data_inventory, inventory_filenames = \
IU.raw2resizednorm_load(raw_dir=img_inventory_dir, img_shape=img_shape)
print("\nx_data_inventory.shape = {0}\n".format(x_data_inventory.shape))
# Explictly assign loaded encoder
encoder = MODEL.encoder
# Encode our data, then flatten to encoding dimensions
# We switch names for simplicity: inventory -> train, query -> test
print("Encoding data and flatten its encoding dimensions...")
if flatten_before_encode: # Flatten the data before encoder prediction
x_data_inventory = IU.flatten_img_data(x_data_inventory)
x_train_kNN = encoder.predict(x_data_inventory)
if flatten_after_encode: # Flatten the data after encoder prediction
x_train_kNN = IU.flatten_img_data(x_train_kNN)
print("\nx_train_kNN.shape = {0}\n".format(x_train_kNN.shape))
x_mean = np.mean(x_train_kNN, axis=0)
x_stds = np.std(x_train_kNN, axis=0)
x_cov = np.cov((x_train_kNN - x_mean).T)
e, v = np.linalg.eig(x_cov)
e_list = e.tolist()
e_list.sort(reverse=True)
plt.clf()
plt.figure(figsize=(1, 1))
plt.bar(np.arange(e.shape[0]), e_list, align='center')
plt.savefig(os.path.join(answer_dir, "1.png"), bbox_inches='tight')
decoder = MODEL.decoder
myev = np.zeros((16))
myev[0] = 0
print(myev)
x = x_mean + np.dot(v, (myev).T).T
print(x)
print(x.shape)
x = np.reshape(x, (2, 2, 4))
print(x.shape)
img = decoder.predict(np.expand_dims(x, axis=0))
PU.save_image(np.squeeze(img, axis=0), os.path.join(answer_dir, "2.png"))
return
# =================================
# Train kNN model
# =================================
print("Performing kNN to locate nearby items to user centroid points...")
EMB = KNearestNeighbours() # initialize embedding kNN class
EMB.compile(n_neighbors=n_neighbors, algorithm=algorithm, metric=metric) # compile kNN model
EMB.fit(x_train_kNN) # fit kNN
# =================================
# Perform kNN on query images
# =================================
# Read items in query folder
print("Reading query images from query folder: {0}".format(query_dir))
# Load query images to memory (resizes when necessary)
x_data_query, query_filenames = \
IU.raw2resizednorm_load(raw_dir=query_dir,
img_shape=img_shape)
n_query = len(x_data_query)
print("\nx_data_query.shape = {0}\n".format(x_data_query.shape))
# Encode query images
if flatten_before_encode: # Flatten the data before encoder prediction
x_data_query = IU.flatten_img_data(x_data_query)
# Perform kNN on each query image
for ind_query in range(n_query):
# Encode query image (and flatten if needed)
newshape = (1,) + x_data_query[ind_query].shape
x_query_i_use = x_data_query[ind_query].reshape(newshape)
x_test_kNN = encoder.predict(x_query_i_use)
query_filename = query_filenames[ind_query]
name, tag = IU.extract_name_tag(query_filename) # extract name and tag
print("({0}/{1}) Performing kNN on query '{2}'...".format(ind_query+1, n_query, name))
if flatten_after_encode: # Flatten the data after encoder prediction
x_test_kNN = IU.flatten_img_data(x_test_kNN)
# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# Compute distances and indices for recommendation
# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
if recommendation_method == 1: # kNN centroid transactions
# Compute centroid point of the query encoding vectors (equal weights)
x_test_kNN_centroid = np.mean(x_test_kNN, axis = 0)
# Find nearest neighbours to centroid point
distances, indices = EMB.predict(np.array([x_test_kNN_centroid]))
elif recommendation_method == 2: # kNN all transactions
# Find k nearest neighbours to all transactions, then flatten the distances and indices
distances, indices = EMB.predict(x_test_kNN)
distances = distances.flatten()
indices = indices.flatten()
# Pick k unique training indices which have the shortest distances any transaction point
indices, distances = find_topk_unique(indices, distances, n_neighbors)
else:
raise Exception("Invalid method for making recommendations")
print(" x_test_kNN.shape = {0}".format(x_test_kNN.shape))
print(" distances = {0}".format(distances))
print(" indices = {0}\n".format(indices))
# =============================================
#
# Output results
#
# =============================================
if output_mode == 1:
result_filename = os.path.join(answer_dir, "result_" + name + ".png")
x_query_plot = x_data_query[ind_query].reshape((-1, img_shape[0], img_shape[1], 3))
x_answer_plot = x_data_inventory[indices].reshape((-1, img_shape[0], img_shape[1], 3))
PU.plot_query_answer(x_query=x_query_plot,
x_answer=x_answer_plot,
filename=result_filename)
elif output_mode == 2:
# Clone answer file to answer folder
# Make k-recommendations and clone most similar inventory images to answer dir
print("Cloning k-recommended inventory images to answer folder '{0}'...".format(answer_dir))
for i, (index, distance) in enumerate(zip(indices, distances)):