论文

EchoNova 发布 Sada-ASR:覆盖八种方言的阿拉伯语语音识别系统

Sada-ASR: A Production Multidialectal Arabic Speech Recognition System

精选理由

一家做阿拉伯语语音客服的公司公开了自己的识别系统,词错率在 120 小时多方言测试里压过 Microsoft、OpenAI、ElevenLabs 的方案。做阿语语音产品的团队可以直接拿他们放的 20 小时测试集和评分代码来测自己的系统。

EchoNova 公布其生产环境阿拉伯语语音智能体所用识别器 Sada-ASR 的技术报告。在 120 小时、覆盖 MSA 与八种方言的测试集上,Sada-ASR 在 12 个系统中取得最低词错率与字符错率:平均词错率 11.69%,低于最强的 Microsoft MAI-Transcribe-2 的 13.51%;八个方言变体平均为 15.52% 对 16.96%。

正文

其优势集中在阿曼(领先 4.30 分)、埃及(3.78 分)和沙特(2.19 分)口音上;叙利亚方言持平,黎巴嫩方言落后(24.96% 对 23.05%),原因主要是它把黎巴嫩词汇转写成 MSA 形式,逐字程度不如对手。在六个公开测试集上,按排行榜当前评分口径其平均词错率为 17.17%,低于 Falcon-ASR(20.92%)、Jais-ASR(22.85%)和 Audar-ASR-V1-Turbo(23.17%),不过后者未声明采用同一评分方式,对比仅供参考。

报告同时披露了数据管线与方法论,并开放 20 小时测试子集与评分代码。对做阿拉伯语呼叫中心或语音智能体的团队而言,方言逐字转写与归一化转写的取舍是文中可复用的细节,测试集也可直接用于自评。

原文 · arXiv: OpenAI

Sada-ASR: A Production Multidialectal Arabic Speech Recognition System

Callers to Arabic customer-service lines speak their own dialect, while speech recognizers are built and evaluated mainly on Modern Standard Arabic (MSA). A recognizer that mishears the dialect breaks the call before anything else can help. We present Sada-ASR, the speech recognizer of EchoNova's production Arabic voice agent. On a 120-hour test set of MSA and eight dialects, it has the lowest word and character error rates of twelve systems, against eleven from Microsoft, ElevenLabs, Google, OpenAI, Speechmatics, IFM, Cohere, NVIDIA and Audar: 11.69% word error rate against 13.51% for the strongest, Microsoft's MAI-Transcribe-2 (a lead of 0.74 points on clips unaffected by that system's fixed number and script conventions), and 15.52% against 16.96% averaged over varieties. Its largest margins are on Omani (4.30 points), Egyptian (3.78) and Saudi (2.19) speech; it is level on Syrian and loses on Lebanese (24.96% against 23.05%), mostly because it writes Lebanese words in MSA forms and is less verbatim. On six public test sets its average word error rate under the leaderboard's current scoring is 17.17%, below every published figure we found, among them Falcon-ASR (20.92%), Jais-ASR (22.85%) and Audar-ASR-V1-Turbo (23.17%), though none of these states that it used that scoring, so the comparison is indicative. We describe the data pipeline and methodology behind Sada-ASR and release a 20-hour subset of the test set with our scoring code.

产业坐标与信源核验

本条情报归属 AITOP「论文」核心追踪栏目。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:Sada-ASR、EchoNova、ASR、语音识别,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。

信源核验说明:本文原始信息发布自一手权威信源「arXiv: OpenAI」(Tier-1 一手发布),具备高确定性参考价值。