Skip to content

feat: increase performance - only calculate embeddings on tensor channels which are not zero #1910

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: develop
Choose a base branch
from
Open

feat: increase performance - only calculate embeddings on tensor channels which are not zero #1910

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 8 additions & 3 deletions pyannote/audio/pipelines/speaker_diarization.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -345,9 +345,14 @@ def iter_waveform_and_mask():
mask_batch = torch.vstack(masks)
# (batch_size, num_frames) torch.Tensor

embedding_batch: np.ndarray = self._embedding(
waveform_batch, masks=mask_batch
)
# remove items from batch which have all zero masks, only compute embeddings for the rest
embedding_batch = np.zeros((mask_batch.shape[0],256), dtype=np.float32)
nonzero_masks = np.where(np.linalg.norm(mask_batch, axis=1) > 0)
if nonzero_masks[0].shape[0] > 0:
embedding_batch_reduced: np.ndarray = self._embedding(
waveform_batch[nonzero_masks], masks=mask_batch[nonzero_masks]
)
embedding_batch[nonzero_masks] = embedding_batch_reduced
# (batch_size, dimension) np.ndarray

embedding_batches.append(embedding_batch)
Expand Down