Lates News

Baidu MuseSteamer audiovisual integrated model announced the completion of industry-first upgrade to achieve multi-person audiovisual integration generation. Its Turbo version, Lite version, Pro version, and full voice version are all open to users, who can experience it through Baidu search for "Baidu MuseSteamer" or log onto the "Imagination" platform. Enterprise users can enjoy high-performance video generation services on the Qianfan platform. According to the introduction, Baidu MuseSteamer is the world's first Chinese audiovisual integrated generation I2V model. It first created the multi-modal latent space planning technology (Latent Multi-Modal Planner), which can autonomously coordinate multiple role identities, emotions, and interactive logic. At the same time, with deep Chinese scene adaptation, it accurately presents Chinese speech details and emotional expressions with over 98% fidelity. In terms of effects, it can achieve movie-quality high-definition video images, realistic environmental sound effects, and synchronized output of natural human voice. Baidu officials also stated that this series of large models has been implemented in various scenarios such as Baidu search, marketing, with pricing as low as 70% of the industry standard (Sina Technology).