近日，清华大学 NLP 实验室、OpenBMB 和面壁智能联合发布并开源了 Ultraeval-Audio，这是一个专为音频模型设计的测评框架。Ultraeval-Audio ...
音频模型评测领域迎来重要突破，由清华大学自然语言处理实验室牵头，联合OpenBMB开源社区与面壁智能团队共同打造的Ultraeval-Audio测评框架正式对外开源。这款专为音频大模型设计的工具包，首次构建了覆盖语音生成、识别、压缩等全场景的标准化评 ...
12月31日，「AI中国」机器之心2025年度评选正式揭晓，梧桐车联科技有限公司（以下简称“TINNOVE梧桐科技”）旗下TTiSoundAI全景声智能座舱音频系统成功入选“最佳大模型产品TOP20”。该产品作为行业首发的由大模型驱动的车载音频方案， ...
IT之家 12 月 23 日消息，阿里通义大模型今日发布了新一代端到端语音交互模型 Fun-Audio-Chat。阿里通义本次开源了 Fun-Audio-Chat 8B，该模型在 OpenAudioBench、MMAU、Speech-ACEBench ...
Powered by Otowahr’s next-generation MEMS speakers, Merry Electronics’ new earbud POC sets a new bar for clarity, speed, and battery life in true-wireless audio SUNNYVALE, CA, UNITED STATES, January 2 ...
OpenAI plans new voice-first GPT models ahead of its screen-free AI device with Jony Ive, focused on natural speech and audio ...
至顶头条 on MSN
Meta推出SAM Audio模型：AI音频分离新突破
Meta发布新款SAM ...
TL;DR: Windows 11 now supports Bluetooth LE Audio with super wideband stereo streaming, enabling improved low-latency, crystal-clear audio and chat for media, gaming, and calls. This update enhances ...
Last year, Apple gave its flagship AirPods Max headphones a subtle refresh when it added new colors and replaced the lightning port with USB-C. But it turns out these small changes had a bigger ...
The Samsung HW-Q990H is the flagship soundbar in the company's 2026 lineup, succeeding last year's HW-Q990F. The HW-Q99H comprises a 7.0.2-channel main bar, 4.0.2-channel rear speakers, and an ...
格隆汇12月23日｜据通义大模型，阿里巴巴发布新一代端到端语音交互模型Fun-Audio-Chat，并已开源8B模型权重、推理代码、Function Call接入示例。
Up against some tough competition. It's undeniably large. Volume control via the streaming apps needs a tweak. Cambridge Audio’s 2021 Evo range of just-add-speaker streaming solutions—complete with ...
一些您可能无法访问的结果已被隐去。
显示无法访问的结果
反馈