Table of Contents
1. What is Luma AI
1.1 About
Luma AI是一家2021年设立的提供3D捕捉和建模技术的美国初创公司，总部位于Palo Alto, California，旨在帮助用户轻松创建高质量的3D内容。他们的技术可以应用于游戏开发、电影制作、虚拟现实等领域，使创作者能够更快速、高效地制作逼真的3D模型和场景。通过深度学习和计算机视觉技术，Luma AI能够快速将现实世界中的物体和场景转换为数字化的3D模型，提高生产效率并降低制作成本。他们的产品包括Luma Scanner和Luma Mesh，分别用于3D扫描和建模，为用户提供丰富的工具和功能来编辑和完善他们的3D模型。
1.2 Team
根据官网资料，目前团队人数32人。
Amit Jain Cofounder & CEO
Amit于2021年共同创立了Luma，并从苹果公司带来了在3D计算机视觉、相机、机器学习、系统工程和深科技产品方面的丰富经验。在加入苹果之前，Amit曾在初创企业工作，并就读于数学和物理学专业。
Alex Yu Cofounder & CTO
Alex 于2021年从加州大学伯克利分校毕业，她在伯克利人工智能研究实验室与Angjoo Kanazawa教授一起进行了关于NeRFs的研究。在BAIR和之前的Adobe期间，Alex致力于推动神经渲染和实时图形领域的发展，提出了Plenoxels（CVPR 2022）、PlenOctrees（ICCV 2021）和pixelNeRF（CVPR 2021）。
1.3 Recruitment
根据官方披露信息，截至本文发稿，Luma AI的招聘信息如下：
招聘职位涵盖了从软件工程到研究科学家的各类高级岗位。从薪资范围图表可以看出，不同职位的薪资范围较为广泛，研究科学家的最高薪资可达$300K，而其它高级工程职位的最高薪资普遍在$250K左右。
1.4 Investor 金主
Andreessen Horowitz (aka a16z) is a venture capital firm that backs bold entrepreneurs building the future through technology. We are stage agnostic. We invest in seed to venture to growth-stage technology companies, across AI, bio + healthcare, consumer, crypto, enterprise, fintech, games, infrastructure, and companies building toward American dynamism. a16z has $42B in assets under management across multiple funds.
2. Luma AI iOS product
2.1 Luma Flythroughs
2.2 Luma AI
3. Luma AI Video to 3D API
Luma的NeRF和网格模型现已在我们的API上提供，使开发人员能够访问3D建模和重建能力。每个场景或物体1美元。
如今，创建3D模型需要花费60至1500美元、2至10周时间，并进行多轮来回沟通。以每个模型一美元的价格，再加上大约30分钟的计算时间，Luma AI 可以为电子商务中整个库存和VFX中的每个预览场景想象出3D模型。
该API期望对象或场景外部向内部拍摄的视频漫游，从2-3层次进行。输出结果是一个可嵌入式交互式3D场景，在传统3D流程中用于构建交互作用的粗糙纹理模型，以及预渲染360度图像和视频。
详见：Luma AI Video to 3D API参考文档。
4. Luma AI Dream Machine
4.1 What is Dream Machine
Luma AI Dream Machine, a next generation video model for creating high quality, realistic shots from text instructions and images using AI.
It is a highly scalable and efficient transformer model trained directly on videos making it capable of generating physically accurate, consistent and eventful shots. Dream Machine is our first step towards building a universal imagination engine and it is available to everyone now!
4.2 Website of Dream Machine
https://lumalabs.ai/dream-machine
4.3 Usage and Pricing
Free 30 generations per month
Standard 120 generations @$29.99/m
Pro 400 generations @$99.99/m
Premier 2,000 generations @$499.99/m
4.4 Interface of Dream Machine
Generations take 120 seconds. Due to high demand requests will be queued.
4.5 Generation methods
- Text only
- Image w or w/ text
- Enhance prompt (Optional)
4.6 Showcase of Dream Machine
5 Dream Machine outlook
The launch of Luma AI’s Dream Machine marks a significant milestone in the rapidly evolving AI video generation market. The overwhelming response to the model’s public beta is a testament to the growing demand for AI-powered tools that can streamline and enhance the video creation process.
What sets Dream Machine apart is its ability to generate high-quality, smooth, and detailed video content from simple text prompts and still images. This level of automation and creative potential is a game-changer for content creators, filmmakers, and businesses alike, as it allows them to explore new avenues of storytelling and visual expression without the traditional constraints of time and resources.
The comparison to OpenAI’s Sora is particularly intriguing, as it highlights the competitive landscape in this space. While Sora has garnered significant attention, Dream Machine’s apparent superiority in certain aspects suggests that the market is ripe for disruption. This competition is likely to drive further innovation and improvements, ultimately benefiting the end-users who seek more powerful and accessible AI video generation tools.
As Luma AI works to address the surge in traffic and improve the user experience, it will be crucial for the company to maintain its focus on quality and reliability. The long-term success of Dream Machine will depend on its ability to consistently deliver high-performance results and establish itself as a trusted and indispensable tool in the creative ecosystem.
Free AI Research Guidebook:
AI Agent Complete Guidebook help gear you up人工智能助手指南
Directly interact with ChatGPT for multi-turn conversations
Input URL as reference material to pass in conversation history, ask multiple questions based on the reference material
Summarize YouTube video summaries, requires enabling subtitles for videos
Summarize and follow up on PDF files
Summarize and follow up on news or web articles
Analyze and ask questions about images
Generate high-quality images
QWEN2本地化部署智体AI AGENT代码
QWEN2本地化部署智能助理AI AGENT代码 local deployment code 构建属于自己的LLM智能助理 Build your own LLM intelligent assistant AI赋能提高你的生产力 AI empowerment improves your productivity