[論文レビュー] SoftDTW-CUDA-Torch: Memory-Efficient GPU-Accelerated Soft Dynamic Time Warping for PyTorch

Ron Shapira Weber, Oren Freifeld|arXiv (Cornell University)|Feb 19, 2026

Time Series Analysis and Forecasting被引用数 0

ひとこと要約

1〜2文の直接的な要約

ABSTRACT

We present softdtw-cuda-torch, an open-source PyTorch library for computing Soft Dynamic Time Warping (SoftDTW) on GPUs. Our implementation addresses three key limitations of existing GPU implementations of SoftDTW: a hard sequence-length cap of 1024, numerical instability in the backward pass for small smoothing parameters, and excessive GPU memory consumption from materializing pairwise distance tensors. We introduce (1) tiled anti-diagonal kernel execution that removes the sequence-length constraint, (2) a log-space back-ward pass that prevents floating-point overflow, and (3) a fused distance-computation mode that eliminates the O(BN M ) intermediate distance tensor, achieving up to 98% memory reduction compared to prior work. The library supports arbitrary sequence lengths, full PyTorch autograd integration, and Soft-DTW Barycenter computation. Code is available at https://github.com/BGU-CS-VIL/sdtw-cuda-torch.

研究の動機と目的

• 研究目的と動機の3〜5点の箇条書き

提案手法

• 提案手法の3〜6点の箇条書き
• 主要技術/式
• ポイント

Figure 1 : Benchmark results for batch size $B=32$ . Top row: Peak GPU memory (MB) as a function of sequence length $L$ (left, $D=128$ ) and feature dimension $D$ (right, $L=256$ ). Bottom row: Wall-clock runtime (ms) for the corresponding configurations. Maghoumi’s implementation is unavailable for

実験結果

リサーチクエスチョン

• RQ1 研究が検証する2〜5の具体的な問い

主な発見

• 主な定量的成果を3〜6点の箇条書き

Figure 2 : SoftDTW Barycenter on synthetic block-wave data.