def __init__(self, model_path, mirror=False):
"""Creates a PoseEngine with given model.
Args:
model_path: String, path to TF-Lite Flatbuffer file.
mirror: Flip keypoints horizontally
Raises:
ValueError: An error occurred when model output is invalid.
"""
BasicEngine.__init__(self, model_path)
self._mirror = mirror
self._input_tensor_shape = self.get_input_tensor_shape()
if (self._input_tensor_shape.size != 4
or self._input_tensor_shape[3] != 3
or self._input_tensor_shape[0] != 1):
raise ValueError(
('Image model should have input shape [1, height, width, 3]!'
' This model has {}.'.format(self._input_tensor_shape)))
_, self.image_height, self.image_width, self.image_depth = \
self.get_input_tensor_shape()
# The API returns all the output tensors flattened and concatenated. We
# have to figure out the boundaries from the tensor shapes & sizes.
offset = 0
self._output_offsets = [0]
for size in self.get_all_output_tensors_sizes():
offset += size
self._output_offsets.append(int(offset))
def __init__(self, model_path, mirror=False):
BasicEngine.__init__(self, model_path)
self._mirror = mirror
self._input_tensor_shape = self.get_input_tensor_shape()
_, self.image_height, self.image_width, self.image_depth = self.get_input_tensor_shape(
)
offset = 0
self._output_offsets = [0]
for size in self.get_all_output_tensors_sizes():
offset += size
self._output_offsets.append(offset)
def __init__(self, model_path):
"""Creates a EmbeddingEngine with given model and labels.
Args:
model_path: String, path to TF-Lite Flatbuffer file.
Raises:
ValueError: An error occurred when model output is invalid.
"""
BasicEngine.__init__(self, model_path)
output_tensors_sizes = self.get_all_output_tensors_sizes()
if output_tensors_sizes.size != 1:
raise ValueError(
('Dectection model should have only 1 output tensor!'
'This model has {}.'.format(output_tensors_sizes.size)))
def __init__(self, model_path):
"""Creates a BasicEngine with given model.
Args:
model_path: String, path to TF-Lite Flatbuffer file.
Raises:
ValueError: An error occurred when the output format of model is invalid.
"""
BasicEngine.__init__(self, model_path)
output_tensors_sizes = self.get_all_output_tensors_sizes()
if output_tensors_sizes.size != 1:
raise ValueError(
('Classification model should have 1 output tensor only!'
'This model has {}.'.format(output_tensors_sizes.size)))
def __init__(self, model_path):
"""Creates a BasicEngine with given model.
Args:
model_path: String, path to TF-Lite Flatbuffer file.
Raises:
ValueError: An error occurred when the output format of model is invalid.
"""
BasicEngine.__init__(self, model_path)
output_tensors_sizes = self.get_all_output_tensors_sizes()
if output_tensors_sizes.size != 1:
raise ValueError(
('Classification model should have 1 output tensor only!'
'This model has {}.'.format(output_tensors_sizes.size)))
def __init__(self, model_path, mirror=False):
"""Creates a PoseEngine with given model.
Args:
model_path: String, path to TF-Lite Flatbuffer file.
mirror: Flip keypoints horizontally
Raises:
ValueError: An error occurred when model output is invalid.
"""
BasicEngine.__init__(self, model_path)
self._mirror = mirror
self._input_tensor_shape = self.get_input_tensor_shape()
if (self._input_tensor_shape.size != 4
or self._input_tensor_shape[3] != 3
or self._input_tensor_shape[0] != 1):
raise ValueError(
('Image model should have input shape [1, height, width, 3]!'
' This model has {}.'.format(self._input_tensor_shape)))
_, self.image_height, self.image_width, self.image_depth = self.get_input_tensor_shape(
)
# The API returns all the output tensors flattened and concatenated. We
# have to figure out the boundaries from the tensor shapes & sizes.
offset = 0
self._output_offsets = [0]
for size in self.get_all_output_tensors_sizes():
offset += int(size)
self._output_offsets.append(offset)
# Auto-detect stride size
def calcStride(h, w, L):
return int(
(2 * h * w) /
(math.sqrt(h**2 + 4 * h * L * w - 2 * h * w + w**2) - h - w))
heatmap_size = self.get_output_tensor_size(4)
print("Heatmap size: ", heatmap_size)
self.stride = calcStride(self.image_height, self.image_width,
heatmap_size)
self.heatmap_size = (self.image_width // self.stride + 1,
self.image_height // self.stride + 1)
print("Stride Guess: ", self.stride, self.heatmap_size)
def __init__(self, model_path):
"""Creates a DetectionEngine with given model.
Args:
model_path: String, path to TF-Lite Flatbuffer file.
Raises:
ValueError: An error occurred when model output is invalid.
"""
BasicEngine.__init__(self, model_path)
output_tensors_sizes = self.get_all_output_tensors_sizes()
if output_tensors_sizes.size != 4:
raise ValueError(
('Dectection model should have 4 output tensors!'
'This model has {}.'.format(output_tensors_sizes.size)))
self._tensor_start_index = [0]
offset = 0
for i in range(3):
offset = offset + output_tensors_sizes[i]
self._tensor_start_index.append(offset)