[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-falcon-asr-arabic-speech-recognition":3,"topics-all":35,"news-related-7fd694d4-7eec-4dd4-8fb7-b6e547da98f4":64},{"id":4,"title":5,"summary":6,"content":7,"original_url":8,"source_id":9,"tags":10,"translations":21,"news_slug":28,"published_at":29,"created_at":30,"modified_at":31,"is_published":32,"publish_type":33,"image_url":14,"view_count":34},"7fd694d4-7eec-4dd4-8fb7-b6e547da98f4","Falcon ASR 开源:1.6B 刷新阿语 WER","阿布扎比 TII 把 Falcon 家族扩进语音,开源 1.6B 参数的 Falcon-ASR,主攻阿拉伯语与海湾方言。按阿拉伯语 ASR 公开榜协议评测,六个测试集平均 WER 20.92%,比榜内已发布最佳低 2.25 个百分点;英语七测试集平均 5.74%。五种语言共用一套权重,免语言标记,支持词级时间戳。","阿布扎比的 Technology Innovation Institute(TII)把 Falcon 家族从文本扩进了语音:开源 1.6B 参数的 Falcon-ASR,主攻阿拉伯语语音识别,重点覆盖海湾方言里的阿联酋方言。模型同时支持英语、法语、西班牙语和葡萄牙语,五种语言共用同一套权重,不需要传语言标记。团队在 Hugging Face 博客发布了完整评测,并开放了在线 Demo。\n\n## 阿语榜:1.6B 拿下平均 WER 20.92%\n\n评测遵循 Open Universal Arabic ASR Leaderboard(ELM Research Center 维护)的协议:六个阿拉伯语测试集、等权平均 WER,同时报 CER。Falcon-ASR 的成绩是平均 WER 20.92%、CER 8.79%,比该榜快照中已发布最佳结果(23.17%)低 2.25 个百分点。官方博客给出的对照:\n\n| 模型 | 参数 | 平均 WER | 平均 CER |\n| --- | --- | --- | --- |\n| Falcon-ASR | 1.6B | 20.92 | 8.79 |\n| Audar-ASR-V1-Turbo | 2.35B | 23.17 | 9.23 |\n| Cohere Transcribe Arabic (07-2026) | 2.0B | 25.87 | 11.80 |\n| omniASR LLM 7B | 7.0B | 28.32 | 12.52 |\n\n对手数字取自该榜 2026-09-30 核验的已发布平均值,不是 TII 自己跑的。\n\n## 方言是主战场:内部阿联酋评测 22.73%\n\n阿拉伯语 ASR 的难点从来不是新闻播报式的标准语,而是方言——方言阿语的转写资源远少于现代标准阿拉伯语(MSA),训练和评测都难。TII 用留出的阿联酋及海湾录音(人工校验转写本)做了内部评测,Falcon-ASR 拿到 WER 22.73%、CER 10.19%,在与 Qwen3-Omni-30B-A3B、Audar、Cohere、Qwen3-ASR 的对照中错误率最低,比次优的 Qwen3-Omni 低 4.07 个百分点。\n\n训练数据里专门混入了背景噪声、重叠语音、音乐、房间混响、电话信道效果以及语速音高变化,且对阿联酋录音做了同样处理——对着真实会议和通话场景去的。\n\n## 英语及其他语言:一套权重,免语言标记\n\n英语侧按 Hugging Face Open ASR Leaderboard 的七个公开测试集评测,平均 WER 5.74%,其中 LibriSpeech clean 1.75%、Earnings-22 11.86%。法语、西语、葡语同权重支持,输出跟随所说语言。另外支持词级时间戳,每个转写词都挂到音频中的位置——字幕和检索类应用直接受益。\n\n架构上,Falcon-ASR 基于 TII 的 Falcon3-Audio 工作,训练方法在其论文(arXiv:2509.07526)中描述。\n\n## 所以呢\n\nFalcon 家族近两个月在阿语栈上连环落子:Falcon-Emirati(方言 LLM)、Falcon OCR Arabic、再到 Falcon-ASR。单看一个模型,1.6B 的体量不算大;但\"方言语料稀缺\"是所有低资源语言的共同处境,TII 的打法——方言优先训练加全条件数据增强,再用公开榜协议自证——对中文方言、东南亚语言这些同样缺转写资源的团队是个可抄的作业。权重已开放,API 访问与原生应用在计划中。\n\n参考:[Introducing Falcon ASR(TII 官方公告)](https:\u002F\u002Fhuggingface.co\u002Fblog\u002Ftiiuae\u002Ffalcon-asr)","https:\u002F\u002Fhuggingface.co\u002Fblog\u002Ftiiuae\u002Ffalcon-asr","c215571f-1ad8-45db-b771-5320c41a3b4e",[11,15,18],{"id":12,"name":13,"slug":13,"description":14,"color":14},"c4455801-0273-477d-9e92-00b4addfa439","asr",null,{"id":16,"name":17,"slug":17,"description":14,"color":14},"7e89b5cc-57db-4f37-bc6d-28919a73931c","model-release",{"id":19,"name":20,"slug":20,"description":14,"color":14},"b9bd9039-fcdb-41a8-b85b-fc1587def2b9","open-source",[22],{"id":23,"lang":24,"title":25,"summary":26,"content":27},"de2e63fa-5c33-448d-b781-3dddd24e5d13","en","Falcon ASR Open-Sourced: 1.6B Sets New Arabic WER Mark","TII open-sources Falcon-ASR: 1.6B Arabic-first, 20.92% avg WER on six test sets (2.25 points under best published); English 5.74%.","Abu Dhabi's Technology Innovation Institute (TII) has extended the Falcon family from text into speech: Falcon-ASR, an open 1.6B-parameter speech recognition model focused on Arabic, with particular attention to the Emirati dialect among Gulf varieties. The same weights also handle English, French, Spanish, and Portuguese — five languages sharing one set of weights with no language flag required. The team published full evaluations in a Hugging Face blog post and opened an online demo.\n\n## Arabic leaderboard: 1.6B posts 20.92% average WER\n\nThe evaluation follows the protocol of the Open Universal Arabic ASR Leaderboard (maintained by the ELM Research Center): six Arabic test sets, equal-weight average WER, with CER also reported. Falcon-ASR scores an average WER of 20.92% and CER of 8.79%, which is 2.25 percentage points below the best published result in the leaderboard snapshot used (23.17%). The comparison from the official blog:\n\n| Model | Parameters | Avg WER | Avg CER |\n| --- | --- | --- | --- |\n| Falcon-ASR | 1.6B | 20.92 | 8.79 |\n| Audar-ASR-V1-Turbo | 2.35B | 23.17 | 9.23 |\n| Cohere Transcribe Arabic (07-2026) | 2.0B | 25.87 | 11.80 |\n| omniASR LLM 7B | 7.0B | 28.32 | 12.52 |\n\nCompetitor figures are the published leaderboard averages verified on 30 September 2026 — not numbers TII ran itself.\n\n## Dialects are the battleground: 22.73% on internal Emirati eval\n\nThe hard part of Arabic ASR has never been formal news-broadcast speech; it is dialect. Dialectal Arabic has far fewer transcribed resources than Modern Standard Arabic, which makes both training and evaluation harder. TII ran an internal evaluation on held-out Emirati and Gulf recordings with human-validated transcripts: Falcon-ASR achieved 22.73% WER and 10.19% CER, the lowest word and character error rates among the systems compared — 4.07 points below Qwen3-Omni, the next best.\n\nTraining data deliberately mixed in background noise, overlapping speech, music, room reverberation, telephony effects, and variations in speed and pitch, with the same treatment applied to Emirati recordings — aimed squarely at real meetings and phone calls.\n\n## English and other languages: one set of weights, no language flag\n\nOn the English side, evaluated across the seven public test sets used by the Hugging Face Open ASR Leaderboard, Falcon-ASR averages 5.74% WER, spanning LibriSpeech clean at 1.75% to Earnings-22 at 11.86%. French, Spanish, and Portuguese are supported by the same weights, with output following the language spoken. Word-level timestamps link each transcribed word to its position in the audio — a direct win for captioning and retrieval applications.\n\nArchitecturally, Falcon-ASR builds on TII's Falcon3-Audio work; the training approach is described in their paper (arXiv:2509.07526).\n\n## So what\n\nThe Falcon family has been stacking up the Arabic stack over the past two months: Falcon-Emirati (dialect LLM), Falcon OCR Arabic, and now Falcon-ASR. A 1.6B model is not large by itself; but \"dialect data is scarce\" is the shared condition of every low-resource language. TII's playbook — dialect-first training plus all-conditions data augmentation, then self-verifying under a public leaderboard protocol — is a copyable recipe for teams working on Chinese dialects or Southeast Asian languages that likewise lack transcription resources. Weights are open; API access and native applications are planned.\n\nReference: [Introducing Falcon ASR (TII official announcement)](https:\u002F\u002Fhuggingface.co\u002Fblog\u002Ftiiuae\u002Ffalcon-asr)","falcon-asr-arabic-speech-recognition","2026-10-11T23:08:05Z","2026-10-11T23:08:27.430913Z","2026-10-11T23:08:27.430926Z",true,"agent",432,[36,46,55],{"slug":37,"tag_slug":38,"title_zh":39,"title_en":40,"intro_zh":41,"intro_en":42,"id":43,"is_active":32,"created_at":44,"modified_at":45},"image-generation","text-to-image","图像生成：从文生图到逐框编辑的模型竞赛","Image Generation: from text-to-image to frame-level editing","文生图已是红海，战场正转向编辑与可控性：FLUX.2\u002F3 用开源和量化把生图搬进消费级显卡，Qwen-Image、Seedream 等国产模型紧咬不放，GPT Image 2 则把推理能力带进像素。本专题追踪图像生成与编辑模型的发布、架构演进、加速量化与商业化数据，看「好图」如何变成「可用的图」。","Text-to-image is a crowded race, and the frontier has moved to editing and control: FLUX.2\u002F3 push open weights and quantization onto consumer GPUs, Qwen-Image and Seedream close in from China, and GPT Image 2 brings reasoning into pixels. This topic tracks model releases, architecture shifts, acceleration work, and the business numbers behind \"pretty\" becoming \"usable\".","52de7c59-7864-4569-8ae7-3d656d26a1a1","2026-11-20T07:20:23.534506Z","2026-11-20T07:20:23.534527Z",{"slug":47,"tag_slug":47,"title_zh":48,"title_en":49,"intro_zh":50,"intro_en":51,"id":52,"is_active":32,"created_at":53,"modified_at":54},"ai-for-science","AI for Science 2026：从 UniPert 到 GPT-Rosalind 的硬核进化","AI for Science 2026: from UniPert to GPT-Rosalind","生命科学、化学材料、物理世界模型——AI 正在从\"语言工具\"变成\"实验伙伴\"。本专题收录 AI 在三大科学方向的关键节点：UniPert 统一基因与化学扰动空间、GPT-Rosalind 端到端生命科学推理、达摩院 AI 智能体 28 小时找到 4 种超导新材料、Anthropic Claude Science 把工作台做成标准品。","From language tool to lab partner — AI is reshaping life sciences, chemistry\u002Fmaterials, and physical world models. This topic covers the key milestones: UniPert unifying genetic-chemical perturbation spaces, GPT-Rosalind's end-to-end life-sciences reasoning, DAMO's AI agent discovering 4 superconducting materials in 28 hours, and Anthropic's Claude Science workbench going mainstream.","988a4300-5fab-41c4-b5d8-63711a2dc757","2026-09-10T01:34:15.296649Z","2026-09-10T01:34:15.296663Z",{"slug":56,"tag_slug":56,"title_zh":57,"title_en":58,"intro_zh":59,"intro_en":60,"id":61,"is_active":32,"created_at":62,"modified_at":63},"h3-series","MiniMax H3 系列：从开源权重到 35 倍吞吐","MiniMax H3 Series: from open weights to 35x throughput","MiniMax H3 自 2026 年 8 月开源以来节奏密集：官方把生成、参考与编辑收回一个模型；ComfyUI 当天压进 RTX 3060；摩尔线程 3 小时完成国产 GPU 适配；fal 后训练版把吞吐拉到 35 倍；FastH3 蒸馏再砍推理成本。本专题持续追踪 H3 的发布—开源—蒸馏—部署全链路。","Since MiniMax open-sourced H3 in August 2026 the pace has been relentless: one unified omni-modal model, same-day ComfyUI support down to an RTX 3060, a 3-hour Day-0 port to Moore Threads GPUs, fal's post-trained H3 Max at 35x throughput, and FastH3 distillation cutting inference cost further. This topic tracks the full H3 chain — release, open weights, distillation, deployment.","83ef0daa-3c31-4cb3-86ed-e5ee58654d5f","2026-09-08T07:33:19.942193Z","2026-09-08T07:33:19.942209Z",{"items":65},[66,71,76,81,86,91],{"id":67,"title":68,"news_slug":69,"published_at":70},"503fc4f0-755d-43f1-89cc-8d86b4cdaed9","7.6M参数土耳其语TTS:WER 1.09%却没人做过盲听","antalia-2-mini-turkish-tts-synthetic-data","2026-11-12T01:40:00+00:00",{"id":72,"title":73,"news_slug":74,"published_at":75},"34bd9f83-04c2-4ffa-8e68-f65a74dfb490","开源语音智能体破50分:Apriel Voice 27B","servicenow-apriel-voice-27b-preview","2026-11-02T13:19:00+00:00",{"id":77,"title":78,"news_slug":79,"published_at":80},"039d2494-af49-4d5d-a89f-e3ca91f60df5","转写没有标准答案:英伟达开源43亿激活语音大模型","nvidia-nemotron-3-5-transcribe","2026-10-29T15:15:27+00:00",{"id":82,"title":83,"news_slug":84,"published_at":85},"35e9f682-7952-4f6a-a8e9-7f8bdd47e791","IBM 语音三件套更新:2B 旗舰 ASR 精度开源第二","granite-speech-5-collection-update","2026-10-22T21:08:48+00:00",{"id":87,"title":88,"news_slug":89,"published_at":90},"d83c4910-6dce-4171-8a44-9f68c9f841af","Qwen-Audio-3.1-ASR开源:转写自带说话人","qwen-audio-3-1-asr-speaker-diarization","2026-10-22T17:13:29+00:00",{"id":92,"title":93,"news_slug":94,"published_at":95},"e7ef7085-3e55-4336-8383-0c1f1dbdfbdd","有道开源 R2T2：不改口流式 ASR 上线","netease-youdao-r2t2-streaming-asr","2026-09-21T09:01:41+00:00"]