IP Library Granted Patent US 12,657,803
Granted Patent B1
US 12,657,803 · App. 18/959,316 · Granted Jun 16, 2026

Cinematic data collection and processing for AI-driven video production

Inventor: Benjamin Geza Affleck-Boldt (West Hollywood, CA)
Assignee: Netflix, Inc.
G06T13/80G06N20/00
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,657,803
App. No.
18/959,316
Granted
Jun 16, 2026
Kind
B1
Abstract

A method enhances AI-driven video production by developing a metadata framework detailing filmmaking techniques, altering variables to demonstrate their impact, and training AI with comprehensive metadata for professional standards. A system includes processors and memory to create a metadata framework, systematically alter filmmaking variables, and train AI with detailed metadata for generating professional video content. A non-transitory computer-readable medium contains instructions for creating a metadata framework, altering filmmaking variables, and training AI with detailed metadata, including spatial Lidar data, for professional video production.

Claims (31)

1 . A method for advanced cinematic data collection and processing for artificial intelligence (AI)-driven video production, the method comprising:

generating a metadata framework that injects detail about filmmaking techniques directly into a learning process of an AI model, including filmmaking elements of camera settings, shot composition, lighting setups, and dynamic scene changes;

systematically altering key variables to teach the AI model an impact of each filmmaking element in the filmmaking elements on video output; and

training the AI model with detailed metadata, including spatial information from Light Detection and Ranging (Lidar) data and filmmaking variables, to generate video content to match technical and/or creative criteria.

2 . The method of claim 1 , further comprising simulating camera movements within the generated video content based on the processed metadata.

3 . The method of claim 1 , further comprising enabling the AI model to replicate or innovate on professional filmmaking techniques in the generated content.

4 . The method of claim 1 , further comprising adjusting lighting within the generated video content in post-production based on the processed metadata.

5 . The method of claim 1 , further comprising generating content narrative having coherence across generated scenes based on the processed metadata.

6 . The method of claim 1 , further comprising employing a feedback mechanism that dynamically refines one or more training parameters of the AI model based on structured feedback mechanisms, enabling iterative improvements in the AI model to produce video content that aligns with the criteria.

7 . The method of claim 1 , further comprising utilizing a simulation-driven learning environment to generate virtual scenes with adjustable parameters, allowing the AI model to learn from hypothetical filmmaking scenarios without the need for continuous acquisition of new real-world video data.

8 . A system for advanced cinematic data collection and processing for AI-driven video production, the system comprising:

one or more processors; and

one or more memories having stored thereon instructions that when executed by the one or more processors, cause the system to:

develop a metadata framework that injects detail about filmmaking techniques directly into a learning process of an AI model, including filmmaking elements of camera settings, shot composition, lighting setups, and dynamic scene changes;

systematically alter key variables to teach the AI model an impact of each filmmaking element in the filmmaking elements on a video output; and

train the AI model with detailed metadata, including spatial information from Lidar Light Detection and Ranging (Lidar) data and filmmaking variables, to generate video content that matches technical and/or creative criteria.

9 . The system of claim 8 , wherein the instructions further cause the system to simulate camera movements within the generated video content based on the processed metadata.

10 . The system of claim 8 , wherein the metadata framework enables the AI model to replicate or innovate on professional filmmaking techniques in the generated content.

11 . The system of claim 8 , wherein the instructions further cause the system to adjust lighting within the generated video content in post-production based on the processed metadata.

12 . The system of claim 8 , wherein the instructions further cause the system to ensure narrative coherence across generated scenes based on the processed metadata.

13 . The system of claim 8 , wherein the instructions further cause the system to employ a feedback mechanism that dynamically refines one or more training parameters of the AI model based on structured feedback mechanisms, enabling iterative improvements in the AI model to produce video content that aligns with the criteria.

14 . A non-transitory computer-readable medium having stored thereon instructions that when executed by one or more processors of a system, cause the system to perform a method for advanced cinematic data collection and processing for AI-driven video production, the method comprising:

developing a metadata framework that injects detail about filmmaking techniques directly into a learning process of an AI model, including filmmaking elements of camera settings, shot composition, lighting setups, and dynamic scene changes;

systematically altering key variables to teach the AI model an impact of each filmmaking element in the filmmaking elements on a video output; and

training the AI model with detailed metadata, including spatial information from Light Detection and Ranging (Lidar) data and filmmaking variables, to generate video content that matches technical and/or creative criteria.

15 . The computer-readable medium of claim 14 , wherein the instructions further cause the system to simulate camera movements within the generated video content based on the processed metadata.

16 . The computer-readable medium of claim 14 , wherein the metadata framework enables the AI model to replicate or innovate on professional filmmaking techniques in the generated content.

17 . The computer-readable medium of claim 14 , wherein the instructions further cause the system to adjust lighting within the generated video content in post-production based on the processed metadata.

18 . The computer-readable medium of claim 14 , wherein the instructions further cause the system to ensure narrative coherence across generated scenes based on the processed metadata.

19 . The computer-readable medium of claim 14 , wherein the instructions further cause the system to dynamically refine one or more training parameters of the AI model based on structured feedback mechanisms, enabling iterative improvements in the AI model to produce video content that aligns with the criteria.

20 . The computer-readable medium of claim 14 , wherein the instructions further cause the system to use a simulation-driven learning environment to generate virtual scenes with adjustable parameters, allowing the AI model to learn from hypothetical filmmaking scenarios without the need for continuous acquisition of new real-world video data.

Assignments (3)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Apr 16, 2026
From: INTERPOSITIVE, LLC
To: NETFLIX, INC.
Reel/Frame 075423/0760 →
CHANGE OF NAME Recorded Nov 24, 2025
From: FIN BONE, LLC
To: INTERPOSITIVE, LLC
Reel/Frame 073322/0142 →
NUNC PRO TUNC ASSIGNMENT Recorded Jul 2, 2025
From: AFFLECK-BOLDT, BENJAMIN GEZA
To: FIN BONE, LLC
Reel/Frame 071595/0698 →
Continuity (1)
Provisional Application 63657756 · Jun 7, 2024
References Cited (99)
US 9681064B1 · Hodges et al. · 2017 [cited by applicant]
US 10645356B1 · Suhy et al. · 2020 [cited by applicant]
US 10771867B1 · Chemolosov · 2020 [cited by examiner]
US 11222667B2 · Ratias · 2022 [cited by applicant]
US 11288864B2 · George et al. · 2022 [cited by applicant]
US 11398255B1 · Mann et al. · 2022 [cited by applicant]
US 11450053B1 · Georgis et al. · 2022 [cited by applicant]
US 11461963B2 · Manivasagam et al. · 2022 [cited by applicant]
US 11570378B2 · Newman · 2023 [cited by applicant]
US 11822719B1 · Chappell, III · 2023 [cited by examiner]
US 11830159B1 · Mann et al. · 2023 [cited by applicant]
US 11928799B2 · Chopra et al. · 2024 [cited by applicant]
US 12033257B1 · Baron · 2024 [cited by examiner]
US 12051205B1 · Deutsch et al. · 2024 [cited by applicant]
US 12133030B2 · Zink et al. · 2024 [cited by applicant]
US 12236517B2 · Bradley et al. · 2025 [cited by applicant]
US 12322036B1 · Affleck-Boldt · 2025 [cited by applicant]
US 20080018667A1 · Cheng et al. · 2008 [cited by applicant]
US 20120102023A1 · Osman et al. · 2012 [cited by applicant]
US 20150286644A1 · Turner et al. · 2015 [cited by applicant]
US 20160071544A1 · Waterston et al. · 2016 [cited by applicant]
US 20160205379A1 · Kurihara · 2016 [cited by applicant]
US 20180124382A1 · Smith et al. · 2018 [cited by applicant]
US 20180136332A1 · Barfield, Jr. et al. · 2018 [cited by applicant]
US 20190304157A1 · Amer · 2019 [cited by examiner]
US 20200082431A1 · Rajasekharan et al. · 2020 [cited by applicant]
US 20200111447A1 · Yaacob et al. · 2020 [cited by applicant]
US 20200125600A1 · Jo · 2020 [cited by examiner]
US 20200364877A1 · Bradski et al. · 2020 [cited by applicant]
US 20210233300A1 · Borer · 2021 [cited by examiner]
US 20210241473A1 · Schmidt et al. · 2021 [cited by applicant]
US 20210241474A1 · Schmidt et al. · 2021 [cited by applicant]
US 20210295148A1 · Gonsalves et al. · 2021 [cited by applicant]
US 20220197306A1 · Cella et al. · 2022 [cited by applicant]
US 20220237410A1 · Wrenninge · 2022 [cited by examiner]
US 20230124190A1 · Chui et al. · 2023 [cited by applicant]
US 20230267706A1 · Djelouah · 2023 [cited by examiner]
US 20230281913A1 · Rematas · 2023 [cited by examiner]
US 20230326114A1 · Baron · 2023 [cited by examiner]
US 20230342481A1 · Nikoghossian et al. · 2023 [cited by applicant]
US 20240005960A1 · Murarka et al. · 2024 [cited by applicant]
US 20240015259A1 · Yu et al. · 2024 [cited by applicant]
US 20240105231A1 · Ratias · 2024 [cited by applicant]
US 20240112394A1 · Neal et al. · 2024 [cited by applicant]
US 20240127794A1 · Seo · 2024 [cited by examiner]
US 20240134926A1 · Tunnicliffe et al. · 2024 [cited by applicant]
US 20240177412A1 · Ranganath et al. · 2024 [cited by applicant]
US 20240290119A1 · Fashandi et al. · 2024 [cited by applicant]
US 20240320918A1 · Amador · 2024 [cited by examiner]
US 20240346731A1 · Graham · 2024 [cited by examiner]
US 20240362897A1 · Klinghoffer et al. · 2024 [cited by applicant]
US 20240394511A1 · Thevenin et al. · 2024 [cited by applicant]
US 20240412542A1 · O'Neill · 2024 [cited by applicant]
US 20240419923A1 · Chollampatt Muhammed Ashraf et al. · 2024 [cited by applicant]
US 20250014606A1 · Wong et al. · 2025 [cited by applicant]
US 20250032945A1 · Mechlowicz · 2025 [cited by applicant]
US 20250148671A1 · Sainz-Nieto · 2025 [cited by applicant]
US 20250190761A1 · Filip et al. · 2025 [cited by applicant]
US 20250204987A1 · Mohareri et al. · 2025 [cited by applicant]
GB 2623644A · 2024 [cited by applicant]
Roush et al., LLM as an Art Director (LaDi): Using LLM's to improve Text-to-Media Generators, 2023 (Year: 2023). [cited by examiner]
Huang et al., Free-Bloom: Zero-Shot Text-to-Video Generator with LLM Director and LDM Animator, 2023 (Year: 2023). [cited by examiner]
Traparic et al., Towards Automatic Content Generation for Immersive Cinema Theater Based on Artificial Intelligence, 2024 (Year: 2024). [cited by examiner]
Lin et al. “VideoDirectorG PT: Consistent Multi-Scene Video Generation via LLM-Guided Planning” 2024. [cited by applicant]
Hong et al., “CogVideo: Large-scale pretraining for text-to-video generation via transformers” 2022. [cited by applicant]
Jiang et al., “Example-driven Virtual Cinematography by Learning Camera Behaviors”, ACM Trans. Graph. 39:4 (2020). [cited by applicant]
Anandraj et al., The impact of AI revolution in transforming film making industry for the digital age, ILIS Journal of Librarianship and Informatics, 6(2):82-91 (Dec. 2023). [cited by applicant]
Azzarelli et al., Reviewing intelligent cinematography: AI research for camera-based video production, arxiv.org, Cornell University Library (May 8, 2024). [cited by applicant]
Bao et al., Vidu: a highly consistent, dynamic and skilled text-to-video generator with diffusion models, arxiv.org, Cornell University Library, 17 pages (May 7, 2024). [cited by applicant]
Chen et al., LiDAR-video driving dataset: learning driving policies effectively, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 5870-5878 (Jun. 18, 2018). [cited by applicant]
De Lima et al., Video-based interactive storytelling using real-time video compositing techniques, Multimed Tools Appl., 77:2333-57 (2018). [cited by applicant]
De Lima et al., Virtual cinematography director for interactive storytelling, Advances in Computer Entertainment Technology, pp. 263070 (Oct. 2009). [cited by applicant]
El Megdani, Visual effects as a filmmaking tool, A research paper submitted as a partial fulfillment of the requirement for getting a bachelor degree in the English department, University Sidi Mohamed Ben Abdellah, 57 p… [cited by applicant]
Ignatov et al., DSLR-quality photos on mobile devices with deep convolutional networks, 2017 IEEE International Conference on Computer Vision, pp. 3297-3305 (Oct. 22, 2017). [cited by applicant]
International Application No. PCT/US2025/032580, International Search Report and Written Opinion, mailed Oct. 9, 2025. [cited by applicant]
International Application No. PCT/US2025/032581, International Search Report and Written Opinion, mailed Oct. 1, 2025. [cited by applicant]
International Application No. PCT/US2025/032582, International Search Report and Written Opinion, mailed Sep. 29, 2025. [cited by applicant]
International Application No. PCT/US2025/032584, International Search Report and Written Opinion, mailed Oct. 9, 2025. [cited by applicant]
International Application No. PCT/US2025/032586, International Search Report and Written Opinion, mailed Sep. 15, 2025. [cited by applicant]
International Application No. PCT/US2025/032594, International Search Report and Written Opinion, mailed Oct. 6, 2025. [cited by applicant]
International Application No. PCT/US2025/032597, International Search Report and Written Opinion, mailed Oct. 6, 2025. [cited by applicant]
International Application No. PCT/US2025/032601, International Search Report and Written Opinion, mailed Oct. 16, 2025. [cited by applicant]
International Application No. PCT/US2025/032603, International Search Report and Written Opinion, mailed Oct. 7, 2025. [cited by applicant]
International Application No. PCT/US2025/032607, International Search Report and Written Opinion, mailed Oct. 13, 2025. [cited by applicant]
International Application No. PCT/US2025/032619, International Search Report and Written Opinion, mailed Sep. 10, 2025. [cited by applicant]
Jiang et al., Cinematographic camera diffusion model, Computer Graphics, 43(2):e15055 (2024). [cited by applicant]
Liu et al., Sora: A review on background, technology, limitations, and opportunities of large vision models, arxiv.org, Cornell University Laboratory, 38 pages (Apr. 17, 2024). [cited by applicant]
Maharaj, Dehancer: the cinematic film look made easy?, accessed online: <https://www.theotivity.com/review/dehancer-the-cinematic-film-look-made-easy-an-honest-review/> (Feb. 24, 2024). [cited by applicant]
Momot, Artificial intelligence in filmmaking process: Future scenarios, Bachelor's thesis, School of Business, JAMK University of Applied Sciences, 46 pages, May 2022. [cited by applicant]
Mumuni et al., A survey of synthetic data augmentation methods in computer vision, arxiv.org, Cornell University Library, 33 pages (Mar. 18, 2024). [cited by applicant]
Nassar, Futuristic scenarios: Utilization of AI technological settings to foster the filmmaking visual creation & mass production, International Design Journal, 14(2):207-12 (Mar. 2024). [cited by applicant]
Paul et al., CamTuner: Reinforcement-learning based system for camera parameter tuning to enhance analytics, arxiv.org, Cornell University Library, 15 pages (Jul. 8, 2021). [cited by applicant]
Pierre et al., CNNs for style transfer of digital to film photography, arxiv.org, 15 pages (Nov. 24, 2024). [cited by applicant]
Wu et al., Automatic camera trajectory control with enhanced immersion for virtual cinematography, IEEE Transactions on Multimedia, 14(8), 16 pages (Aug. 2015). [cited by applicant]
Wu et al., The secret of immersion: actor driven camera movement generation for auto- cinematography, arxiv.org, Cornell University Library (May 23, 2023). [cited by applicant]
Yu et al., Enabling automatic cinematography with reinforcement learning, 2022 IEEE 5th International Conference on Multimedia Information Processing and Retrieval (MIPR), pp. 103-108 (Aug. 2022). [cited by applicant]
Zhu et al., MovieFactory: Automatic movie creation from text using large generative models for language images, arxiv.org, Cornell University Library, 7 pages (Jun. 12, 2023). [cited by applicant]
International Application No. PCT/US2025/032611, International Search Report and Written Opinion, mailed Nov. 14, 2025. [cited by applicant]
Galvane, Automatic Cinematography and Editing in Virtual Environments, Dissertation, Universite Grenoble Alpes (2015). [cited by applicant]