|TensorFlow 1 version||View source on GitHub|
Reconstructs a signal from a framed representation.
tf.signal.overlap_and_add( signal, frame_step, name=None )
Adds potentially overlapping frames of a signal with shape
[..., frames, frame_length], offsetting subsequent frames by
The resulting tensor has shape
[..., output_size] where
output_size = (frames - 1) * frame_step + frame_length
signal: A [..., frames, frame_length]
Tensor. All dimensions may be unknown, and rank must be at least 2.
frame_step: An integer or scalar
Tensordenoting overlap offsets. Must be less than or equal to
name: An optional name for the operation.
Tensor with shape
[..., output_size] containing the overlap-added
signal's inner-most two dimensions.
signal's rank is less than 2, or
frame_stepis not a scalar integer.