def get_options():
return {
'architectures': list(ModelBuilder.ARCHITECTURES.keys()),
'datasets': list(ModelBuilder.DATASETS.keys()),
'losses': list(ModelBuilder.LOSSES.keys()),
'optimizers': list(ModelBuilder.OPTIMIZERS.keys()),
'clusters': ClusterRegistry.keys(),
}
def run():
try:
if not flask.request.is_json:
response = flask.Response()
response.status_code = 400
response.status = 'Error: Not JSON'
return response
request_data = flask.request.get_json()
directory = request_data['url']
settings = Settings()
settings.__dict__.update(request_data)
# if not settings.exists():
# settings.save()
# else:
# settings.load()
cluster = ClusterRegistry.get(settings.cluster)
model = ModelBuilder.create(
settings.architecture,
settings.dataset,
settings.loss,
settings.optimizer,
settings.metrics,
settings.epochs,
settings.patience,
)
split = model.split(settings.split)
results = main(directory, cluster, settings.clusterArgs, split)
results = flask.jsonify(results)
return results
except TrainingIncompleteException:
response = flask.Response()
response.status_code = 400
response.status = 'Error: Training incomplete'
return response
except ModelStateMissingError:
response = flask.Response()
response.status_code = 400
response.status = 'Error: Model state missing'
return response
except BadModelSettings:
response = flask.Response()
response.status_code = 400
response.status = 'Error: Incompatible architecture/dataset'
return response
except ClusterRegistryNameError:
response = flask.Response()
response.status_code = 400
response.status = 'Error: Bad cluster name'
return response
except:
print_exc()
response = flask.Response()
response.status_code = 500
response.status = 'Error: Unknown'
return response
def cluster_algorithms():
results = list()
for name, algorithm in ClusterRegistry.items():
parameters = function_signature(algorithm.run)
if 'images' in parameters:
del parameters['images']
results.append({
'name': name,
'parameters': parameters,
})
return flask.jsonify(results)
def cluster():
try:
if not flask.request.is_json:
response = flask.Response()
response.status_code = 400
response.status = 'Error: Not JSON'
return response
settings = flask.request.get_json()
images = ImageDirectory(settings['directory']).jpeg(False)
cluster = ClusterRegistry.get(settings['cluster'])
results = cluster.run_cached(images, **settings['args']).urls()
results = flask.jsonify(results)
return results
except ClusterRegistryNameError:
response = flask.Response()
response.status_code = 400
response.status = 'Error: Unknown clustering algorithm'
return response
except:
print_exc()
response = flask.Response()
response.status_code = 500
response.status = 'Error: Unknown'
return response
class HistogramCluster(ClusterStrategy):
"""
"""
def run(
self,
images: List[Url],
hue_bins: int = 180,
saturation_bins: int = 256,
value_bins: int = 256,
bandwidth: Optional[float] = None,
) -> ClusterResults:
"""
Clusters images.
"""
c = list()
for i, img in enumerate(images):
print("HISTOGRAM: %i / %i" % (i, len(images)))
hh = HsvHistogram(img)
histogram = HsvHistogram.scale(hh.hsv(hue_bins, saturation_bins, value_bins), hh.size())
c.append(histogram)
d = vstack(c)
print('CLUSTER: Mean Shift')
cluster = MeanShift(bandwidth=bandwidth).fit_predict(d).tolist()
return ClusterResults(images, cluster)
ClusterRegistry.add('histogram', HistogramCluster())
if similarity_metric == SimilarityMetric.INVERSE_DISTANCE:
inverse_distance = 0
for k in good:
inverse_distance += 1 - k.distance
if len(good) > 0:
matrix[i][j] = inverse_distance / len(good)
matrix[j][i] = inverse_distance / len(good)
elif similarity_metric == SimilarityMetric.COUNT:
matrix[i][j] = len(good)
matrix[j][i] = len(good)
else:
if similarity_metric == SimilarityMetric.INVERSE_DISTANCE:
matrix[i][i] = 1
elif similarity_metric == SimilarityMetric.COUNT:
matrix[i][i] = nfeatures
print('CLUSTER: AffinityPropagation')
cluster = AffinityPropagation(
damping=damping,
max_iter=max_iter,
convergence_iter=convergence_iter,
affinity=affinity.value,
random_state=0,
).fit_predict(matrix).tolist()
return ClusterResults(images, cluster)
ClusterRegistry.add('sift', SiftCluster(Similarity1()))
ClusterRegistry.add('sift2', SiftCluster(Similarity2()))
ClusterRegistry.add('sift3', SiftCluster(Similarity3()))
ClusterRegistry.add('sift4', SiftCluster2())
damping=damping,
max_iter=max_iter,
convergence_iter=convergence_iter,
affinity=affinity,
descriptor_matcher=descriptor_matcher,
)
results2 = HistogramCluster().run_cached(
images,
hue_bins=hue_bins,
saturation_bins=saturation_bins,
value_bins=value_bins,
bandwidth=bandwidth,
)
labels1 = results1.labels()
labels2 = results2.labels()
k1 = results1.k()
cluster = [
self.combine(c1, c2, k1) for c1, c2 in zip(labels1, labels2)
]
self.remove_empty_clusters(cluster)
return ClusterResults(images, cluster)
ClusterRegistry.add('hybrid1a', HybridCluster(Similarity1()))
ClusterRegistry.add('hybrid1b', HybridCluster(Similarity2()))
ClusterRegistry.add('hybrid1c', HybridCluster(Similarity3()))
ClusterRegistry.add('hybrid2a', HybridCluster2(Similarity1()))
ClusterRegistry.add('hybrid2b', HybridCluster2(Similarity2()))
ClusterRegistry.add('hybrid2c', HybridCluster2(Similarity3()))
ClusterRegistry.add('hybrid3', HybridCluster3())