Workflow
语音合成
icon
Search documents
智谱开源GLM-TTS工业级语音合成系统
Mei Ri Jing Ji Xin Wen· 2025-12-11 05:22
每经AI快讯,12月11日,据智谱公众号消息,智谱正式上线并开源GLMTTS工业级语音合成系统。只 需3秒语音样本,GLMTTS即可学习说话人的音色和说话习惯。在通用朗读、情感配音、教育评测、电 子书、有声客服等场景中,实现自然流畅、贴近真人的语音。 (文章来源:每日经济新闻) ...
老年人权益保护要稳定“大池子”筑牢“小池子”
第一财经· 2025-12-03 00:59
Core Viewpoint - The article emphasizes the need for stronger legal protections and societal efforts to safeguard the rights of the elderly, particularly against fraud and exploitation in the context of pension funds and health-related scams [2][4][6]. Group 1: Legal Framework and Enforcement - The Supreme Court has released six typical cases of fraud affecting the elderly, highlighting the serious violations of their rights, including scams related to pension funds and health products [2][4]. - Recent improvements in legal frameworks, such as the revision of the Elderly Rights Protection Law and the issuance of guidelines by multiple government bodies, aim to reduce risks of rights violations for the elderly [4][5]. - The Supreme Court noted that many perpetrators lack a clear understanding of the legal implications of their actions, indicating a need for clearer rules and stricter enforcement of existing laws [5]. Group 2: Evolving Fraud Tactics - Fraud targeting the elderly has shifted from traditional methods to more sophisticated online tactics, including the use of AI technologies to impersonate relatives and create convincing scams [5]. - Many elderly victims have low awareness of evolving fraud techniques, which complicates the recovery of lost funds and emphasizes the need for improved awareness and timely reporting of scams [5][6]. Group 3: Societal Responsibility - Building an elderly-friendly society requires collective efforts from various sectors, including legal support and community involvement, to stabilize the broader social security system and protect individual rights [6].
清华大学与巨人网络联合首创多方言语音合成框架;腾讯IEG CDD总经理刘智鹏加入游戏科学丨游戏早参
Mei Ri Jing Ji Xin Wen· 2025-10-15 23:23
Group 1: Tsinghua University and Giant Network Collaboration - Tsinghua University and Giant Network AI Lab have jointly developed the DiaMoE-TTS framework for multi-dialect speech synthesis, which is fully open-source [1] - The DiaMoE-TTS framework aims to address the challenges in dialect TTS, which has been a "gray area" in the industry due to the reliance on proprietary data [1] - The framework is designed to be comparable to industrial-grade dialect TTS models, utilizing a unified IPA expression system based on linguistic expertise [1] Group 2: Giant Network's Competitive Advantage - Giant Network's differentiation in the AI model competition lies in its focus on technology and binding to specific scenarios, particularly in gaming [2] - Unlike comprehensive AI firms like Baidu and iFlytek, Giant Network's speech synthesis technology is tailored to meet the localization needs of games [2] - The company benefits from natural advantages in scene implementation and cash flow support compared to pure AI startups [2] Group 3: Fire Feather Game's New Release - Fire Feather Game announced the closed beta test for its self-developed casual management mobile game "Dream Diary," starting on October 20 and running until October 30 [3] - The game targets the casual management genre, aligning with market demand for lightweight entertainment products [3] - The closed beta test will help the company gather player feedback and optimize payment design ahead of the official launch [3] Group 4: Talent Movement in the Gaming Industry - Liu Zhipeng, General Manager of Tencent's Interactive Entertainment Group (IEG) CDD, has joined Game Science, supported by Tencent [4] - Liu's expertise in content ecosystem building and project management is expected to enhance Game Science's capabilities in premium development and IP long-term operation [4] - The transition is seen as a strategic move that may lead to an increase in Game Science's market valuation due to expected synergies with Tencent's resources [4]
昆仑万维:Mureka V7.5模型正式上线 AI音乐创作水平再迎新高度
Core Insights - Kunlun Wanwei officially launched the Mureka V7.5 model on August 15, enhancing the performance of Chinese song interpretation significantly [2] - The Mureka V7.5 model demonstrates a deep understanding of various Chinese music styles, allowing for accurate emotional and artistic expression in generated music [2] - The company also introduced MoE-TTS, a novel speech synthesis framework that combines pre-trained large language model capabilities with specialized speech expert modules [3] Group 1 - Mureka V7.5 has improved the timbre and performance techniques of Chinese songs, as well as the articulation and emotional expression [2] - The model's deep accumulation of knowledge regarding Chinese music diversity enables it to convey unique artistic essence and emotional nuances [2] - The ASR technology has been optimized to enhance the authenticity and emotional depth of vocal performances in generated music [2] Group 2 - MoE-TTS innovatively integrates pre-trained large language model text capabilities with speech expert modules, ensuring independent optimization of each modality [3] - The release of MoE-TTS provides a reproducible open descriptive TTS solution for academia and demonstrates the potential of decoupled modalities and knowledge freezing in speech synthesis [3] - Future plans for MoE-TTS include integration into the Mureka-Speech platform, offering customizable descriptive speech synthesis capabilities for global developers and creators [3]