# Fully Differentiable Neural Forced Alignment via Soft Dynamic Programming

arXiv: https://arxiv.org/abs/2606.25460 · HTML: https://mlsanity.com/paper/2606.25460/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 26 references that are in the map.
- First posted on arXiv 24 Jun 2026.

## Builds on (most cited first, 10 of 10)

- [Representation Learning with Contrastive Predictive Coding](https://mlsanity.com/paper/1807.03748/) (2018)
- [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://mlsanity.com/paper/2006.11477/) (2020)
- [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://mlsanity.com/paper/2106.07447/) (2021)
- [Scaling Speech Technology to 1,000+ Languages](https://mlsanity.com/paper/2305.13516/) (2023)
- [WhisperX: Time-Accurate Speech Transcription of Long-Form Audio](https://mlsanity.com/paper/2303.00747/) (2023)
- [Soft-DTW: a Differentiable Loss Function for Time-Series](https://mlsanity.com/paper/1703.01541/) (2017)
- [Canary-1B-v2 & Parakeet-TDT-0.6B-v3: Efficient and High-Performance Models for Multilingual ASR and AST](https://mlsanity.com/paper/2509.14128/) (2025)
- [Self-Supervised Contrastive Learning for Unsupervised Phoneme Segmentation](https://mlsanity.com/paper/2007.13465/) (2020)
- [Tradition or Innovation: A Comparison of Modern ASR Methods for Forced Alignment](https://mlsanity.com/paper/2406.19363/) (2024)
- [Phoneme Boundary Detection using Learnable Segmental Features](https://mlsanity.com/paper/2002.04992/) (2020)

## Built on by (most cited first, 1 of 1)

- [FA-Bench: A Benchmark for Phone- and Word-Level Timestamp Accuracy in Forced Alignment and ASR on Clean and Noisy Speech](https://mlsanity.com/paper/2609.32396/) (2026)

## Abstract (arXiv)

Recent advances in sequence modeling have significantly improved ASR systems, bringing them close to human-level recognition accuracy and enhancing robustness across diverse acoustic conditions and languages. In contrast, Forced Alignment has not experienced comparable progress, and traditional HMM-GMM frameworks remain widely adopted and highly competitive. To address this gap, we propose an end-to-end, fully differentiable neural architecture specifically designed for phoneme alignment. The model consists of an encoder that processes the input signal and a decoder that produces alignment decisions. The encoder is structured into two complementary branches: one dedicated to phoneme identity verification and the other to phoneme boundary detection. The decoder is implemented as a trainable module based on differentiable soft dynamic programming. The entire system is optimized end-to-end using a novel contrastive loss that encourages clear separation between steady-state phoneme regions and transition boundaries. The proposed approach outperforms the current state of the art in phoneme alignment on hand-annotated English benchmarks, achieves strong word-level generalization results, and demonstrates generalization on unseen languages.