据新浪科技报道，近日有关字节跳动已推出颠覆性视频模型Boximator的传闻在业界引发关注。据称，Boximator具备通过文本精确操控视频中人物或物体动作的独特功能，被视作可能挑战Sora在文生视频赛道的领先地位。然而，字节跳动官方对此进行了澄清，表示Boximator目前仅是一个视频生成领域的技术研究项目，尚未发展到可以作为完整产品推出的阶段。
字节跳动的相关人士指出，尽管Boximator在视频内容的文本控制方面取得了一定进展，但其在画面质量、保真度以及视频时长等关键指标上，与国际领先的视频生成模型相比仍有显著差距。因此，该技术目前并不具备商业化落地的条件，仍需进一步的研发和完善。
这一回应打破了市场上关于字节跳动已推出与Sora竞争产品的猜测，同时也显示出字节跳动在技术创新上的谨慎态度。公司将继续致力于视频生成技术的研究，以期在未来能推出更为成熟、先进的产品，满足用户对高质量视频内容的需求。
英语如下：
Title: “ByteDance Denies Launch of Chinese Version of Sora: Boximator Remains in R&D, Far from Commercialization”
Keywords: ByteDance, Boximator, Video Generation
News Content:
ByteDance has refuted recent reports that it has launched the groundbreaking video model Boximator, which reportedly has the unique capability of precisely controlling the actions of characters or objects in videos through text, potentially challenging Sora’s dominance in the text-to-video space. As per Sina Tech, the news sparked interest in the industry. However, the company clarified that Boximator is currently an ongoing research project in the domain of video generation and is not yet ready for commercial release.
ByteDance insiders explained that while Boximator has made progress in textual control of video content, it still lags significantly behind leading video generation models in crucial aspects like image quality, fidelity, and video duration. Consequently, the technology is not yet suitable for commercialization and requires further development and refinement.
This response dispels speculation that ByteDance has introduced a product to compete with Sora and underscores the company’s cautious approach to technological innovation. The company remains committed to researching video generation technology, aiming to eventually推出 more mature and advanced products that cater to users’ demands for high-quality video content.
【来源】https://finance.sina.com.cn/tech/shenji/2024-02-20/doc-inairynk7158099.shtml
Views: 1