en https://localhost:3000/favicon.ico 语音杂谈 公众号“语音杂谈”更新 - 使用 Huginn 制作 https://localhost:3000 Sun, 03 May 2026 20:19:39 -0700 Sun, 03 May 2026 20:19:39 -0700 60 MeanAudio：单步生成！100 倍加速！首个 MeanFlow 文本转音频模型重磅开源 http://mp.weixin.qq.com/s?__biz=MzI4OTQyNzA0Ng%3D%3D&mid=2247544976&idx=2&sn=9be632d0d04c1d2747e4e5e8bce5dbaf Mon, 13 Apr 2026 03:00:45 -0700 633087 语音/音频处理学术速递[4.13] http://mp.weixin.qq.com/s?__biz=MzI4OTQyNzA0Ng%3D%3D&mid=2247544976&idx=3&sn=aadf8a8d0daa937374abf382f95ac949 Mon, 13 Apr 2026 03:00:45 -0700 633086 【征稿启事】第十五届中文口语语言处理国际研讨会 (ISCSLP 2026) 邀您相聚马来西亚槟城！ http://mp.weixin.qq.com/s?__biz=MzI4OTQyNzA0Ng%3D%3D&mid=2247544976&idx=1&sn=dc1d49eec285190f0da066086a7d3969 Mon, 13 Apr 2026 03:00:44 -0700 633073 语音/音频处理学术速递[3.30] http://mp.weixin.qq.com/s?__biz=MzI4OTQyNzA0Ng%3D%3D&mid=2247544680&idx=3&sn=481e008444f526046928f2e85b3151c9 Mon, 30 Mar 2026 03:08:39 -0700 630630 Suno 5.5 发布 允许用户克隆声音演唱 AI 生成歌曲 http://mp.weixin.qq.com/s?__biz=MzI4OTQyNzA0Ng%3D%3D&mid=2247544680&idx=2&sn=1284d747138a494153c7b44722c8f1dc Mon, 30 Mar 2026 03:08:39 -0700 630628 ICASSP 2026｜MeanVC：基于平均流的轻量级流式零样本语音转换 http://mp.weixin.qq.com/s?__biz=MzI4OTQyNzA0Ng%3D%3D&mid=2247544680&idx=1&sn=be550df1d0c4764a623a4e5619f39e72 Mon, 30 Mar 2026 03:08:38 -0700 630627 基于WeNet的关键词识别WeKws - WeNet keyword spotting-3-其他模型构造 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496939&idx=4&sn=4c3d2acbdfdc05231161c62acf39168a Fri, 29 Jul 2022 19:00:39 -0700 629374 基于WeNet的关键词识别WeKws - WeNet keyword spotting-2-TCN模型forward http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496939&idx=3&sn=5413692e8dbce0dac27cf24ffdebcc23 Fri, 29 Jul 2022 19:00:39 -0700 629373 基于WeNet的关键词识别WeKws - WeNet keyword spotting-1-TCN模型构造 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496939&idx=2&sn=ca269d2d65c22a3fe7d09bb940224870 Fri, 29 Jul 2022 19:00:39 -0700 629372 INTERSPEECH 2022论文解读｜Paraformer: 高识别率、高计算效率的单轮非自回归端到端语音识别模型 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496939&idx=1&sn=5aedc86899d5064fcb27583bf5c00180 Fri, 29 Jul 2022 19:00:38 -0700 629362 极速上手新一代 Kaldi 服务端框架 sherpa http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496610&idx=2&sn=702413199010b8b2fff18c109b6eacbf Wed, 20 Jul 2022 03:30:47 -0700 629361 端到端语音识别应用基于前缀树的热词技术 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496610&idx=1&sn=f61290bb301a4cae89ccf0e40462b301 Wed, 20 Jul 2022 03:30:46 -0700 629360 论文分享|腾讯音乐天琴实验室两篇论文入选国际会议INTERSPEECH http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496482&idx=2&sn=8251ff59810cd7db14a5d62ee06edff0 Sat, 16 Jul 2022 03:30:47 -0700 629359 IEEE TPAMI | 火山语音提出多源迁移高斯回归模型，效果超越多个SOTA方法 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496482&idx=1&sn=b7c7abc9be451939e252f3ca4f46f2dd Sat, 16 Jul 2022 03:30:46 -0700 629358 论文推介：语音指令识别中的最小序列混淆错误准则 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496436&idx=2&sn=4c37d9c3083c1b4f606acc6491028fd4 Thu, 14 Jul 2022 03:31:23 -0700 629357 【语音之家】AI产业沙龙—智能语音技术在美团的应用实践 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496436&idx=1&sn=40ec5d167c75a93276ca51afe2e2de73 Thu, 14 Jul 2022 03:31:23 -0700 629338 字节跳动 AI Lab 总监李航：语言模型的过去、现在和未来 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496373&idx=2&sn=281711deeff01720fb00eaf50344fd8f Tue, 12 Jul 2022 03:31:40 -0700 629337 新一代 Kaldi 中基于量化的蒸馏实验 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496373&idx=1&sn=83341a63d8450889c0439a6b92644625 Tue, 12 Jul 2022 03:31:40 -0700 629336 AIWIN 2022丨文本语音驱动数字人表情口型竞赛 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496327&idx=2&sn=d9320e99fe006766b6ecb6582c6d0e2c Sat, 09 Jul 2022 04:02:16 -0700 629334 论文分享丨NPU-ASLP实验室将携14篇论文参加语音旗舰会议INTERSPEECH2022 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496327&idx=1&sn=3934c9f011515101c700ae336e417155 Sat, 09 Jul 2022 04:02:16 -0700 629333 论文分享丨基于随机循环一致性准则的语音信息分解工作获 Odyssey 最佳学生论文奖 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496241&idx=2&sn=8c9ed0a13aff52be3df983c4f32fee02 Thu, 07 Jul 2022 04:01:33 -0700 629332 如何基于新一代 Kaldi 框架快速搭建服务端 ASR 系统 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496241&idx=1&sn=d412722ba58c56014d54f0a621bb03ac Thu, 07 Jul 2022 04:01:33 -0700 629314 论文分享 | THUHCSI人机语音交互实验室论文入选多媒体领域顶级国际会议ACM MM http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496194&idx=3&sn=38a26a9d6e56624f4f635ab5c86300d6 Tue, 05 Jul 2022 04:00:26 -0700 629312 全球首个《AI对话系统分级定义》发布，这下语音助手有了强弱之分 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496194&idx=2&sn=99c348c4b93aba26bfae2e1c2a605d60 Tue, 05 Jul 2022 04:00:25 -0700 629310 【语音之家公开课】语音分离的神经网络方法 Speech Separation with Neural Network http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496194&idx=1&sn=ca6de9eba781847e2dcb7610c181b166 Tue, 05 Jul 2022 04:00:24 -0700 629309 Reworked Conformer http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496130&idx=2&sn=657cb8bfa4e2610f60410c7ecb94c798 Sat, 02 Jul 2022 04:00:29 -0700 629308 如何打造真人化高表现力的语音合成系统 http://mp.weixin.qq.com/s?__biz=MzU4MTA0NDE5NQ%3D%3D&mid=2247496130&idx=1&sn=b2e8e7a4723d41de80c4871f913777ef Sat, 02 Jul 2022 04:00:29 -0700 629307 ICASSP 2026｜Phys-NVAS：基于视觉语言模型物理先验挖掘与3D声学环境建模的新视角声学合成 http://mp.weixin.qq.com/s?__biz=MzI4OTQyNzA0Ng%3D%3D&mid=2247544660&idx=1&sn=f4dfc47f37775070c428a8de20216888 Sat, 28 Mar 2026 03:00:30 -0700 627673