GstAnalytics.Tensor

Fields

Name	Type	Access	Description
data	Gst.Buffer	r/w	Gst.Buffer holding tensor data
data_type	GstAnalytics.TensorDataType	r/w	GstAnalytics.TensorDataType of tensor data
dims	[int]	r/w	number of tensor dimensions
dims_order	GstAnalytics.TensorDimOrder	r/w	Indicate tensor elements layout in memory.
id	int	r/w	semantically identify the contents of the tensor
layout	GstAnalytics.TensorLayout	r/w	Indicate tensor layout
num_dims	int	r/w	number of tensor dimensions

Methods

class	alloc (num_dims)
class	new_simple (id, data_type, data, dims_order, dims)
	copy ()
	free ()
	get_dims ()

Details

class GstAnalytics.Tensor

Hold tensor data

New in version 1.26.

classmethod alloc(num_dims)

Parameters:

num_dims (int) – Number of dimension of the tensors

Returns:

tensor allocated

Return type:

GstAnalytics.Tensor

Allocate a tensor with num_dims dimensions.

New in version 1.26.

classmethod new_simple(id, data_type, data, dims_order, dims)

Parameters:

• id (int) – semantically identify the contents of the tensor

• data_type (GstAnalytics.TensorDataType) – GstAnalytics.TensorDataType of tensor data

• data (Gst.Buffer) – Gst.Buffer holding tensor data

• dims_order (GstAnalytics.TensorDimOrder) – Indicate tensor dimension indexing order

• dims ([int]) – tensor dimensions. Value of 0 mean the dimension is dynamic.

Returns:

A newly allocated GstAnalytics.Tensor

Return type:

GstAnalytics.Tensor

Allocates a new GstAnalytics.Tensor of dims_order ROW_MAJOR or COLUMN_MAJOR and with an interleaved layout

New in version 1.26.

copy()

Returns:

a new GstAnalytics.Tensor

Return type:

GstAnalytics.Tensor or None

Create a copy of self.

New in version 1.26.

free()

Free tensor

New in version 1.26.

get_dims()

Returns:

The dims array form the tensor

Return type:

[int]

Gets the dimensions of the tensor.

New in version 1.26.