Draft:Act-Two (AI motion capture)
 | Review waiting, please be patient.
This may take 8 weeks or more, since drafts are reviewed in no specific order. There are 2,955 pending submissions waiting for review.
Where to get help
How to improve a draft
You can also browse Wikipedia:Featured articles and Wikipedia:Good articles to find examples of Wikipedia's best writing on topics similar to your proposed article. Improving your odds of a speedy review To improve your odds of a faster review, tag your draft with relevant WikiProject tags using the button below. This will let reviewers know a new draft has been submitted in their area of interest. For instance, if you wrote about a female astronomer, you would want to add the Biography, Astronomy, and Women scientists tags. Editor resources
Reviewer tools
|
Submission declined on 19 September 2025 by Protobowladdict (talk).
Where to get help
How to improve a draft
You can also browse Wikipedia:Featured articles and Wikipedia:Good articles to find examples of Wikipedia's best writing on topics similar to your proposed article. Improving your odds of a speedy review To improve your odds of a faster review, tag your draft with relevant WikiProject tags using the button below. This will let reviewers know a new draft has been submitted in their area of interest. For instance, if you wrote about a female astronomer, you would want to add the Biography, Astronomy, and Women scientists tags. Editor resources
This draft has been resubmitted and is currently awaiting re-review. |  |
Comment: In accordance with the Wikimedia Foundation's Terms of Use, I disclose that I have been paid by my employer for my contributions to this article. Itsjuliamartins (talk) 19:21, 19 September 2025 (UTC)
Act-Two is an AI-powered motion capture and character animation model developed by Runway AI Inc., released in July 2025. The model captures head, face, body, and hand movements to create character animations from video inputs.
History
[edit]Act-Two succeeded Runway's Act-One model, which was released in 2024[1][2] as part of the Gen-3 Alpha video generation system.[3] Act-One captured facial animations and expression transfer using single-camera setups,[4] but was limited to head and facial movement capture without body and hand tracking capabilities.[5][6]
Act-Two was released in July 2025, with initial access provided to Enterprise customers and Creative Partners, followed by paid subscribers.[7][8] The model uses webcam or smartphone recordings as input, without requiring motion capture suits or specialized equipment.[9][10]
Technical Details
[edit]Architecture
[edit]Act-Two uses computer vision and machine learning algorithms to analyze video inputs and extract motion data including facial expressions, body posture, hand gestures, and micro-expressions.[11] The model transfers movements from driving performance videos to character references.[12] Act-Two operates within Runway's Gen-4 video generation system through Chat Mode and Tool Mode interfaces.[13]
Capabilities
[edit]Act-Two captures head, face, body, and hand movements including facial expressions, eye movements, finger positions, and body posture.[7] The model synchronizes speech and mouth movements between the driving performance and animated character, and can modify character voices.[14][15]
Act-Two processes various character types including humans, cartoon characters, creatures, and abstract representations.[13][15]
Specifications
[edit]The model generates video at 24 frames per second in various aspect ratios including 16:9 (1280×720), 9:16 (720×1280), and 1:1 (960×960).[16] Runway recommends sequences under 30 seconds. Act-Two requires a driving performance video and a character reference (image or video) as inputs.[13][6]
Usage and Applications
[edit]Media Production
[edit]Act-Two is used in film and television production for character performances and animated sequences, particularly for pre-visualization and rapid prototyping.[7] Production companies employ the model to test scenes before filming without traditional motion capture facilities.[17]
The model is also used for creating short films and web series. Content creators use the model to produce animated narratives in shorter timeframes than traditional animation workflows.[10][18]
Interactive Media
[edit]Game developers use Act-Two for character animation and NPC development in both independent and commercial game production.[12][19] The model generates character animations for cutscenes, in-game dialogue sequences, and interactive character moments.[20]
VR developers apply Act-Two for creating character interactions in immersive environments, including virtual actors for VR experiences and interactive narratives.[21] The technology is integrated with game engines including Unity and Unreal Engine for real-time and pre-rendered character animations.[22]
Commercial and Educational Content
[edit]Marketing agencies employ Act-Two for branded content and advertising campaigns featuring animated spokespersons or product demonstrations.[19][23]
Educational institutions use Act-Two for instructional videos, creating animated instructors for online courses and training materials.[12]
Reception
[edit]Trade publications have documented Act-Two's use in professional workflows without traditional motion capture studio equipment.[7][24] The model includes full-body motion capture, distinguishing it from facial-only solutions like LivePortrait.[25]
Limitations
[edit]Results vary based on alignment between driving performance videos and character references, with mismatched perspectives or low-resolution inputs affecting output quality.[26] Performance varies with non-human characters that differ significantly from human facial structure.[25] Act-Two includes content moderation systems to prevent generation of unauthorized public figure content, with policy violations potentially resulting in task rejection or account restrictions.[27]
Availability
[edit]Act-Two requires Runway's Standard plan or higher, with usage based on credit consumption per video duration.[28] Free-tier users cannot access the model. The feature is available through Runway's web platform, API,[26] and mobile applications.[29]
References
[edit]- ^ Wiggers, Kyle (2024-10-23). "This Week in AI: The AI agents cometh". TechCrunch. Retrieved 2025-09-22.
- ^ Franzen, Carl (2024-10-22). "'This is a game changer': Runway releases new AI facial expression motion capture feature Act-One". VentureBeat. Archived from the original on 2025-08-02. Retrieved 2025-09-22.
- ^ "Runway Research | Introducing Act-One". runwayml.com. Retrieved 2025-09-19.
- ^ read, Chris McKay • 1 min (2024-10-22). "Runway's Act-One Lets You Animate AI-Generated Characters with a Video". Maginative. Retrieved 2025-09-22.
{{cite web}}: CS1 maint: numeric names: authors list (link) - ^ "Creating with Act-One on Gen-3 Alpha and Turbo". Runway. Retrieved 2025-09-19.
- ^ a b Land 🌐, V. P. "Runway's Act-Two is here". VP Land. Retrieved 2025-09-22.
- ^ a b c d "Runway Launches Next-Gen Motion Capture AI Model With Aim to Revolutionize MoCap Industry | No Film School". nofilmschool.com. Retrieved 2025-09-19.
- ^ DeepNewz (2025-07-15). "Runway Unveils Act-Two AI for Full-Body Motion Capture | DeepNewz AI Modeling". DeepNewz. Retrieved 2025-09-22.
- ^ "Introducing Runway Act Two". 2025-07-16. Retrieved 2025-09-19.
- ^ a b Brent Le Blanc (2025-08-08). Runway Act 2 AI Mocap Tutorial — Turn Any Video into Animation. Retrieved 2025-09-22 – via YouTube.
- ^ Ai, Simple With (2025-07-16). "Introducing Runway Act-Two: Revolutionizing Motion Capture for Animation". Simple with Ai. Retrieved 2025-09-19.
- ^ a b c "Runway Act Two: The AI-Powered Motion Capture and Performance Transfer Tool - Ai Adoption Agency". 2025-07-27. Retrieved 2025-09-19.
- ^ a b c "Creating with Act-Two". Runway. Retrieved 2025-09-19.
- ^ "Creating Multi-Character Dialogues with Act-Two". Runway. Retrieved 2025-09-19.
- ^ a b OpenArt (2025-09-19). Runway Act-Two is the EASIEST Way to Create AMAZING Characters!. Retrieved 2025-09-22 – via YouTube.
- ^ "Models". Runway API. Retrieved 2025-09-19.
- ^ Land 🌐, V. P. "Runway's Act-Two is here". VP Land. Retrieved 2025-09-19.
- ^ "Introducing Runway Act Two". 2025-07-16. Retrieved 2025-09-19.
- ^ a b Ai, Simple With (2025-07-16). "Introducing Runway Act-Two: Revolutionizing Motion Capture for Animation". Simple with Ai. Retrieved 2025-09-22.
- ^ Roberts, Emily. "AI in gaming will soon transform everythings". Remocapp. Retrieved 2025-09-19.
- ^ "Introducing Runway Act Two". 2025-07-16. Retrieved 2025-09-19.
- ^ "AI in Game Development: How Studios Are Using It Today". Udonis Mobile Marketing Agency. 2025-08-29. Retrieved 2025-09-19.
- ^ "Runway Act-Two: El nuevo salto de la IA creativa que deberías conocer si trabajas en publicidad o contenidos - Madrid Content School". www.madridcontent.school (in Spanish). Retrieved 2025-09-19.
- ^ AI, Escapism. "Runway Act-Two: Next-Gen Motion Capture for Effortless Character Animation". Escapism AI. Retrieved 2025-09-19.
- ^ a b "Runway Act Two vs. Live Portrait - Atlabs AI". www.atlabs.ai. Retrieved 2025-09-19.
- ^ a b "Runway/Act_two - CometAPI - All AI Models in One API". 2025-09-14. Retrieved 2025-09-19.
- ^ "Runway Research | Introducing Act-One". runwayml.com. Retrieved 2025-09-19.
- ^ "Pricing | Runway". runwayml.com. Retrieved 2025-09-19.
- ^ "Runway | Tools for human imagination". runwayml.com. Retrieved 2025-09-19.
- in-depth (not just passing mentions about the subject)
- reliable
- secondary
- independent of the subject
Make sure you add references that meet these criteria before resubmitting. Learn about mistakes to avoid when addressing this issue. If no additional references exist, the subject is not suitable for Wikipedia.