AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

DASpeech: Directed Acyclic Transformer for Fast and High-quality Speech-to-Speech Translation

Qingkai Fang, Yan Zhou, Yang Feng

NeurIPS 2023 · Architectures and efficiency › RNNs and normalization · approach: Non-autoregressive translation · 6 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
expect-path training instead of best-path training (lambda=0.5, Lookahead) (ablation)DASpeech · CVSS-C Fr-En testASR-BLEU: 24.45 -> 24.71Improved
expect-path training instead of best-path training (lambda=0.5, Joint-Viterbi) (ablation)DASpeech · CVSS-C Fr-En testASR-BLEU: 24.84 -> 25.03Improved
expect-path training instead of best-path training (lambda=1.0, Lookahead) (ablation)DASpeech · CVSS-C Fr-En testASR-BLEU: 24.18 -> 24.41Improved
expect-path training instead of best-path training (lambda=1.0, Joint-Viterbi) (ablation)DASpeech · CVSS-C Fr-En testASR-BLEU: 24.46 -> 24.80Improved
end-to-end expect-path finetuning instead of cascading DAT and FastSpeech 2 (lambda=0.5, Lookahead)DASpeech · CVSS-C Fr-En testASR-BLEU: 22.19 -> 24.71Improved
end-to-end finetuning instead of cascade (lambda=1.0, Joint-Viterbi)DASpeech · CVSS-C Fr-En testASR-BLEU: 23.20 -> 24.80Improved
smaller graph size lambda=0.5 instead of 1.0 (Joint-Viterbi) (ablation)DASpeech · CVSS-C Fr-En testASR-BLEU: 24.80 -> 25.03Improved
larger graph size lambda=1.0 instead of 0.5 (Joint-Viterbi) (ablation)DAT + FastSpeech 2 cascade · CVSS-C Fr-En testASR-BLEU: 22.80 -> 23.20Improved
Joint-Viterbi decoding instead of Lookahead (lambda=0.5) (ablation)DASpeech · CVSS-C Fr-En testASR-BLEU and speedup: 24.71 -> 25.03; 18.53x -> 16.29xMixed
non-autoregressive DASpeech without knowledge distillation instead of TranSpeech with NPDDASpeech · CVSS-C Fr-En testASR-BLEU: 19.05 -> 25.03Improved
DASpeech instead of Translatotron 2 in multilingual X-En settingDASpeech · CVSS-C multilingual X-En testASR-BLEU: 8.74 -> 7.43Hurt
DASpeech instead of S2UT in multilingual X-En settingDASpeech · CVSS-C multilingual X-En testASR-BLEU: 5.15 -> 7.43Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy