/pytorch-softdtw-cuda-with_lengths

Fast CUDA implementation of (differentiable) soft dynamic time warping for PyTorch, with lengths specified for samples in batch.

Primary LanguagePythonMIT LicenseMIT

Soft DTW for PyTorch in CUDA (with lengths support)

Forked from Maghoumi's implementation, but added the function to support specifying lengths for each sample in batch.

License

This project is licensed under the MIT License.