开启时会模糊预览图，关闭后正常显示
Veo3 和 Sora2 发布之后，AI 视频模型的发展趋势已经非常清晰，就是是音视频融合。 今天，国内的终于跟上节奏，火山引擎正式发布了Seedance 1.5 pro。 它采用了业界领先的音视频联合生成方案。 它在音画同步、方言演绎、运镜控制和叙事表达上展现了惊人的潜力。 这是我做的一组相声和方言的实测。
除了人声，环境音效（SFX）与画面的同步率也非常重要，可以简化以前复杂的 AI 音效工作流。特别是在游戏场景、电影特效场景等有高频音效需求的场景里。 这个模型不仅懂画面声音，还懂专业的摄影术语，在运镜上表现得更加专业，连续性更好。 人物表演方面也更加细腻真实。 这是几个测试示例，实测抽卡成功率 50%。
As holidays approach, i bet some tech guy is gonna post pic of their laptop w/ "grind doesn't stop" or something.
and then ghost of christmas future will decapitate them.
$1K at birth is a great idea until you realize it’s all going to go directly into sports gambling and robinhood the moment the kid is old enough
“you’re being too cynical, they still have 18 years for it to grow until-“ the market is smart and there are no laws anymore. the game of every consumer app is to get users as young as possible
«They don't have Trumpf, Zeiss, Cymer» Yeah? But that's just engineering? It's sad to see westerners reduced to this condition. The idea seems to be that we've got some legacy tech from German dwarves, and can now coast on it. Why exactly is Zeiss more special than BMW or Kuka?
We're in a race. It's not USA vs China but humans and AGIs vs ape power centralization. @deepseek_ai stan #1, 2023–Deep Time «C’est la guerre.» ®1
RT @youcc: 北美留学党必看！🔥 老俞（俞敏洪）当年刚到美国，走进校园餐厅听到学生聊天，瞬间震惊：“这不就是托福听力真题里的对话吗？！” 事实证明：想练地道口语，托福听力真题就是最好的“活教材”！那些校园场景、日常聊天，超级贴合北美生活，刷熟了出门零尴尬！… htt…
独立开发者 ｜ 个人IP教练 ｜ 帮助新手在X上完成早期成长｜ 公众号：PandaTalk8
RT @MaximillianNYC: We recently added reply support to @yutori_ai Scouts. Just react to each report & Scout will figure it out! https://t.c…
Co-founder & Chief Scientist @yutori_ai. Prev: Senior Director leading FAIR Embodied AI @MetaAI and Professor @GeorgiaTech.