Workflow
Google DeepMind
icon
Search documents
Veo 3.1 - Create longer, seamless shots
Google DeepMind· 2025-10-15 15:56
With "Extend," you can create longer videos, even lasting for a minute or more, that connect to and continue the action from your original clip. Each video is generated based on the final second of your previous clip, making it most useful for creating a longer establishing shot. Try it today in Flow at flow.google. Learn more: https://blog.google/technology/ai/veo-updates-flow ____ Subscribe to our channel / @googledeepmind Find us on X / googledeepmind Follow us on Instagram / googledeepmind Add us on Lin ...
Veo 3.1 and more artistic control in Flow
Google DeepMind· 2025-10-15 15:56
Product Updates - Veo 3.1 introduces richer audio, more narrative control, and enhanced realism [1] - Veo 3.1 builds on Veo 3, with stronger prompt adherence and improved audiovisual quality for image-to-video conversion [1] - New capabilities are introduced, bringing audio to existing capabilities for the first time [1] Technology - Veo 3.1 is state-of-the-art [1]
Veo 3.1 - Designed to empower creatives
Google DeepMind· 2025-10-15 15:56
We're giving creators more artistic control with increased support for audio across all features. We’re also bringing audio to existing capabilities like “Ingredients to Video,” “Frames to Video” and “Extend.” We’re also introducing Veo 3.1, which brings richer audio, more narrative control, and enhanced realism that captures true-to-life textures. Veo 3.1 is state-of-the-art and builds on Veo 3, with stronger prompt adherence and improved audiovisual quality when turning images into videos. Try it today at ...
Beyond phishing: Cyber threats in the age of AI with Four Flynn (pt. 1)
Google DeepMind· 2025-10-09 18:27
Social engineering, cyberattacks, and the fog of war - all topics covered in this interview with the VP of Security and Privacy at Google DeepMind. Hannah Fry and Four Flynn take us behind the scenes of Operation Aurora, the monumental 2009 attack on Google that forever changed the landscape of cybersecurity. They discuss the defender's dilemma, the constant battle between attackers and defenders in the digital world, and how AI can potentially help mitigate some of the most complex vulnerabilities. As Hann ...
From sketches to prototype: Designing with generative AI
Google DeepMind· 2025-10-01 16:00
AI & Design Collaboration - Google DeepMind 与艺术家合作,探索 AI 的应用,并利用艺术家的创造力来影响 AI 技术的发展 [1] - Google 使用 Ross Lovegrove 的草图训练图像生成模型,以产生基于其独特设计语言的新想法 [2] - AI 生成了数百个椅子的迭代和排列方案 [2] - Gemini 被用于优化设计,从多个角度可视化椅子设计,并通过先进的 3D 打印工艺实现 [3] Project Outcome & Future Implications - 该项目展示了 AI 如何帮助人们以新的维度进行思考 [4] - AI 可以为设计过程带来独特而非凡的东西 [4] - 该项目预示着 AI 在设计领域的未来应用 [4]
Gemini Robotics 1.5: Using agentic capabilities
Google DeepMind· 2025-09-25 15:54
[Music] Hi, Aloha. Can you use my location and sort the objects into the correct compost, recycling, and trash bins. To perform the sorting task based on San Francisco waste guidelines, I will use the green bin for compost, the blue bin for recycling, and the black bin for trash.Here we have an agentic layers. Both control the robot and have access to the internet. To sort the trash, they had to look up the rules and look at the objects on the table, think about where each object should go, matching the rul ...
Gemini Robotics 1.5: Learning across embodiments
Google DeepMind· 2025-09-25 15:54
This is usually how we train our robots is actually someone teleoperates the robot. So they have to physically move the robot. And right now what he's doing is he's training the robot how to do a task.So traditionally people will train a single model per robot. In Geminina Robotics 1.5%, we're actually using a single model across multiple robots. So, here's an example.We're hanging things from that workbench. One of the key things in 1.5% is that now all of our other robots can actually do the same task. Th ...
Gemini Robotics 1.5: Enabling robots to plan, think and use tools to solve complex tasks
Google DeepMind· 2025-09-25 15:54
Earlier this year, we brought Gemini's multimodal understanding to the physical world with Gemini robotics, allowing robots to behave in interactive, dextrous, and general ways. Previously, robots could complete one task per instruction. >> Previous Gemini Robotics version has been tested over and over and over again to put this banana into the bowl.This is a very simple task. >> Today, we've reached a new milestone. We're introducing Gemini Robotics 1.5%, a new family of models to power the next generation ...
Gemini Robotics 1.5: Thinking while acting
Google DeepMind· 2025-09-25 15:53
[Music] Previous Gemini Robotics version has been tested over and over and over again to put this banana into the bowl. This is a very simple task. I'm going to switch the model here and make the task slightly more challenging.[Music] So this is Gemini Robotics 1.5%. Aloha. Clear sort these fruits into color matching plates.>> Sure, I can certainly help you do that. Put the green fruit into the green plate. >> We enable it to think.It can perceive the environment, different colors of the object, different c ...
Google DeepMind researchers react to Nano Banana demos 🍌
Google DeepMind· 2025-09-24 17:26
I think the fact that people surprise us with a model we built is the best idea. So, so this is like a demo with nano banana hooked up into I think it's an studio demo. It's hooked onto a canvas and you can like drag these isometric shapes around.Oh, and you're so cool. I mean, we often thought of like Nano Banana as a single tool, as a single thing, but now actually this becomes more part of a pipeline. Wait, San Francisco. They merged San Francisco, New York halfway.What. Oh, no way. Oh, wow.Is that the B ...