The Storyteller Company develops tools for automating filmmaking and interactive streaming experiences. We leverage machine learning and deepfake technology to change actors' appearances, motion capture to rig actors with prebuilt models, volumetric photogrammetry to composite actors into scenes, and more.
We're pioneers of interactive storytelling, and our tools can also be leveraged by audiences in real time to change settings, spawn enemies, and create virtual on-screen cheering and heckling from deepfaked cartoon characters, celebrities, and more.
You can teleport yourself directly into the action with volumetric capture. Instead of needing a camera crew, the world is compositied and moves around you!
We support a variety of off-the-shelf depth sensors, including Microsoft's Azure Kinect line and Intel RealSense.
We're planning on building support for a camera-only solution using cinema-quality lenses.
Viewers can interact with streamers in real time and change the narrative flow. They can voice themselves through text to speech, spawn enemies, change the setting, and more. This unlocks entirely new types of content for streamers: highly interactive improv theater, wildly fun games of dungeons and dragons, and more. The viewers are in control.
Streamers can even interact with games played within games.
Wire up rigged models to motion capture sensors to inhabit a new body. We support wearables and computer vision-based solutions.
You can even pull off-the-shelf models from Models Resource and other popular websites.
We offer lots of flexibility for Virtual YouTubers and Streamers.
Our users can create compelling content easily with the power of deep fakes. This can be done in our engine or on our website. Users can download content to create media for sharing and remixing.
Vo.codes is our deep fake platform with a friendly web interface. It lets users create novel audio and video content and easily remix it for distribution on social media.
Vo.codes is also integrated into our streaming engine as a compelling community engagement feature.
Our "vo.codes" text to speech engine combines several of the latest machine learning architectures to generate audio from well-known characters and celebrities.
We also have a voice conversion system that changes a speaker's voice to a target speaker (pending release). Collectively these tools allow our community to create compelling new audio content that has proven wildly popular.
Our users can also create deepfake videos using a few of the most popular algorithms: Wav2Lip and first-order-model.
Vo.codes makes it easy for users to create funny and engaging content. It's especially useful for folks that don't have beefy GPUs or that lack the expertise to deal with code.
The community that creates with us is also heavily involved in data curation and annotation for our ML models. To that end, we have a variety of tools that make gathering data and labeling it easy.
Users have stepped up and taken an active role in building data sets that extend our capabilities and reach.
As we build more features that directly involve measurement — mocap-based posture estimation, camera movement, timeline editing, etc. — we'll collect data that will help us train models to automate even more creative tasks.
Tool of the Week: AI voice generator | Vo.codes [...] project from Brandon Thomas is a window into the future [...]. Play with it with a number of celebrity voices, including Judi Dench, Neil DeGrasse Tyson, and Bill Gates.
無料でビル・ゲイツやアーノルド・シュワルツネッガーなど有名人に好きな台詞をしゃべらせることができる「Vocodes」レビュー
("Vocodes" allows users to use celebrities such as Bill Gates and Arnold Schwarzenegger to speak their favorite lines for free.)
It’s simply a piece of AI software called Vocodes. The tool, which I can best describe as a deepfake generator, can mimic the voices of a slew of politicians and celebrities including Donald Trump, Barack Obama, Bryan Cranston, Danny Devito, and a dozen more.
Have you ever wanted David Attenborough to narrate your audiobook? Judi Dench to read your shopping list? Gilbert Gottfried to... well... some things are better left unsaid.
These are recent videos created using the "vo.codes" deepfake toolset, which is both a standalone website featuring deepfake tools, and a stream-integrated audience interaction platform.
Our community's videos collectively have tens of millions of views.
In Hollywood, it used to be that there was only so much room at the top, and not enough budget to go around. Not every starry-eyed creative could become a Steven Spielberg or a Patty Jenkins.
Tomorrow's content creation is faster, easier, and cheaper. Because of that, it will undoutably be much more accessible. It will look less like filming on a Hollywood back lot or editing in Adobe Premiere and more like ML auto assist with knobs, sliders, and high level control planes. Despite the simplicity, users will still have control when they want it, and the results will have the look and feel of a well-polished Scorsese film.
With advancements in ML, computer vision, spatial computing, and semantic understanding, you'll have to go out of your way to make a bad film or live stream.
We have a plan to lay the cornerstones for our angle of attack.
We're cultivating a community of creators: streamers, YouTubers, and game designers. They will use our tools and guide their evolution.
Users can contribute data to build shared models, insert themselves into each others' streams, and remix content to create new memes or narratives.
We're taking the cost and difficulty out of camera work, video editing, and more. Our tools will better enable creatives by leveraging automation. A virtual camera requires no camera crew.
Over time, our models will learn what makes for compelling plot and pacing and will offer assistance on narrative structure, shot composition, and timeline construction.
Hollywood and traditional media are quadratic in cost, time, and complexity. The next wave of technology includes fixes for dozens of these inefficiencies, and transforms human labor back into human creativity. Imagine if you could make a film in a weekend.
We go beyond traditional "full stack". Computer Vision, Machine Learning, Game Engine programming, as well as traditional backend and frontend roles. We're at the forefront of where "spatial computing" meets automation and entertainment.
Anyone can create highly polished content. Most have simply never been afforded the opportunity.
