Tag: #speech-ai11 results
Alibaba's Qwen team launched Qwen3.5-Omni, a full-modal AI model. It supports speech recognition in 113 languages and speech generation in 36 languages. The model sets 215 state-of-the-art results across multimodal benchmarks.
Alibaba has launched CosyVoice Studio, described as China’s first integrated AI voice platform. The platform combines listening, speaking, and content creation while incorporating semantic understanding into voice capabilities.
The update highlights the integration of xAI's Grok into the Cursor code editor. It also mentions developments regarding GPT-5.6 and updates to the ChatGPT voice interface.
CONCORD is a privacy-aware A2A framework for proactive speech-based AI assistants that captures only owner speech via real-time verification, producing one-sided transcripts. It recovers missing context through spatio-temporal resolution, gap detection, and minimal relationship-aware A2A queries, avoiding hallucinations. Evaluations show 91.4% gap detection recall, 96% relationship classification, and 97% privacy TNR.
SpeakFlow, built for Z.AI hackathon, is an AI coach evaluating spoken responses in Practice and Presentation modes across 11 languages. Uses Web Speech API and GLM 5.1 for real-time scoring and reports. Live demo and GitHub code available.
Amsterdam-based Reson8 raised €5M in pre-seed funding to scale its customizable speech AI platform. It offers high-precision, industry-specific recognition tailored for European languages, challenging US-dominated solutions. Plans include expanding models to over 20 languages.
Ghanaian developer Emmanuel Quartey’s temporary hearing loss revealed how poorly everyday interactions can support people with sensory impairments. The article explores his effort to redesign assistive AI around the needs and experiences of African users.
Google Translate celebrates its 20th anniversary with a new AI-powered pronunciation coach. The feature provides real-time feedback by listening to user speech and pinpointing exact errors. This enhances language practice directly in the app.
Google launched an AI-powered pronunciation practice feature in Translate for Android, available for English, Spanish, and Hindi in the US and India. Users can speak translations for instant AI feedback on accuracy, including phonetic spellings. The app supports 250+ languages and has 1 billion monthly users translating 1 trillion words.