今日，有媒体报道称字节跳动推出了一款名为Boximator的视频生成模型，该模型能够通过文本精准控制生成视频中人物或物体的动作。然而，字节跳动方面迅速作出回应，澄清Boximator目前只是一个技术研究项目，尚未达到可以作为产品推出的成熟阶段。
据字节跳动相关人士表示，尽管Boximator在视频生成技术上展现出了一定的创新性，但在画面质量、保真率和视频时长等方面，与国外领先的视频生成模型相比，仍然存在较大的差距。这意味着，字节跳动在推出中文版Sora之前，还需要进一步研究和开发，以确保其技术能够达到市场预期。
此外，字节跳动还强调，目前Boximator还无法作为一个完善的产品落地，公司将继续投入研发资源，以缩小与国外同行的技术差距。
英文标题：ByteDance Clarifies Non-Launch of Chinese Version of Sora, Video Generating Technology Still in Research Phase
英文关键词：ByteDance, Video Generation, Technology R&D
英文新闻内容：
Today, reports surfaced that ByteDance had launched a new video generation model called Boximator, which reportedly allows for precise control of movements in generated videos through text. However, ByteDance quickly issued a response clarifying that Boximator is currently a research project and is not yet ready to be commercialized as a product.
According to a spokesperson from ByteDance, while Boximator demonstrates certain innovative capabilities in video generation, it still lags behind leading foreign models in terms of video quality, fidelity, and duration. This suggests that ByteDance will need to invest further in research and development before launching a Chinese version of Sora.
Moreover, ByteDance emphasized that Boximator is not yet ready to be launched as a complete product and that the company will continue to allocate resources to research in order to bridge the technological gap with foreign competitors.
【来源】https://finance.sina.com.cn/tech/shenji/2024-02-20/doc-inairynk7158099.shtml
Views: 0