DCT
8:25-cv-02362
Artificial Intelligence Industry Association Inc v. Topaz Labs LLC
Key Events
Complaint
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: Artificial Intelligence Industry Association, Inc. (Florida)
- Defendant: Topaz Labs, LLC (Texas)
- Plaintiff’s Counsel: The Strategic Legal Group, PLLC
- Case Identification: 8:25-cv-02362, M.D. Fla., 09/03/2025
- Venue Allegations: Plaintiff alleges venue is proper because Defendant committed acts of infringement in the district, including sales to customers with operations in the district such as NASA's Kennedy Space Center. The complaint notes that establishing venue based on a "regular and established place of business" may require further discovery.
- Core Dispute: Plaintiff alleges that Defendant’s artificial intelligence-powered software products for image and video enhancement infringe three U.S. patents related to the creation of three-dimensional video, three-dimensional images, and image color equalization.
- Technical Context: The dispute is in the field of computational photography and videography, where software algorithms are used to enhance or transform digital media, a market with significant commercial and consumer applications.
- Key Procedural History: The complaint states that prior to filing the lawsuit, Plaintiff sent Defendant a formal demand letter identifying the Asserted Patents, which may establish a date for knowledge in the context of willfulness allegations.
Case Timeline
| Date | Event |
|---|---|
| 2009-07-31 | U.S. Patent No. 8,508,580 Priority Date |
| 2010-11-03 | U.S. Patent No. 9,185,388 Priority Date |
| 2012-10-04 | U.S. Patent No. 8,965,121 Priority Date |
| 2013-08-13 | U.S. Patent No. 8,508,580 Issued |
| 2015-02-24 | U.S. Patent No. 8,965,121 Issued |
| 2015-11-10 | U.S. Patent No. 9,185,388 Issued |
| 2025-06-10 | Date of product investigation cited in complaint |
| 2025-09-03 | Complaint Filing Date |
II. Technology and Patent(s)-in-Suit Analysis
U.S. Patent No. 9,185,388 - "Methods, Systems, and Computer Program Products for Creating Three-Dimensional Video Sequences"
The Invention Explained
- Problem Addressed: The patent describes the challenge of creating three-dimensional (stereoscopic) video content, noting that while dual-lens cameras exist, it is desirable to create 3D video from more common, conventional single-lens 2D video sources (’388 Patent, col. 1:22-43, 63-67).
- The Patented Solution: The invention proposes a method that processes a standard 2D video sequence by selecting a target frame and then analyzing two different subsets of surrounding frames. A first subset with a large spatial displacement (a "large stereo baseline") is used to extract depth data for static background objects, while a second subset with a small spatial displacement (a "smaller stereo baseline") is used to calculate the depth of moving objects. This combined depth information is then used to generate a 3D video frame ('388 Patent, Abstract; col. 2:7-24).
- Technical Importance: This approach provides a method to generate 3D video from ubiquitous 2D video footage, potentially expanding the availability of 3D content without requiring specialized capture hardware ('388 Patent, col. 1:63-67).
Key Claims at a Glance
- The complaint asserts infringement of at least independent Claim 1 (Compl. ¶8).
- Claim 1 requires a method comprising the steps of:
- receiving a two-dimensional video sequence including a plurality of frames;
- selecting a target frame from among the plurality of frames;
- selecting a first subset of frames (N) from the plurality of frames that are associated with the target frame and representative of a large stereo displacement;
- extracting depth data of static objects utilizing the first subset of frames;
- selecting a second subset of frames (n) from the plurality of frames that are associated with the target frame, representative of a stereo displacement substantially different than N;
- utilizing the second subset of frames to calculate depth of moving objects; and
- generating a three-dimensional video frame based on the depth data.
U.S. Patent No. 8,508,580 - "Methods, Systems, and Computer-Readable Storage Media for Creating Three-Dimensional (3D) Images of a Scene"
The Invention Explained
- Problem Addressed: The patent's background section notes the widespread use of conventional 2D digital capture devices and states that it is desirable to provide methods for using such devices to create 3D images (’580 Patent, col. 1:46-62).
- The Patented Solution: The invention discloses a multi-step process for converting a plurality of 2D images of a scene into a single 3D image. The process includes receiving images, determining their attributes to select a suitable pair for stereoscopic conversion, and then performing a series of processing steps including registration, rectification, transformation, and color correction. A key part of the solution involves calculating parallax values and, if they do not meet certain criteria, altering stereoscopic characteristics to create an adjusted, final 3D image ('580 Patent, Abstract; col. 2:5-13).
- Technical Importance: The patented method allows for the creation of 3D stereoscopic images from a series of standard 2D still images, such as those taken by a conventional digital camera, thereby obviating the need for specialized 3D camera hardware ('580 Patent, col. 1:56-62).
Key Claims at a Glance
- The complaint asserts infringement of at least independent Claim 1 (Compl. ¶17).
- Claim 1 requires a method comprising the steps of:
- receiving a plurality of captured images;
- determining matching image features among the plurality of captured images;
- determining a transformation to align at least one of the matched image features;
- applying the transformation to a first captured image to create a plurality of transformed images;
- defining a transformed image pair using one of transformed and captured images;
- calculating parallax values of the matched image features of the transformed image pair;
- determining whether the parallax values do not meet a predetermined criteria;
- in response to the determination, altering a stereoscopic characteristic of the transformed image pair to create an adjusted image pair; and
- defining a three-dimensional image based on the adjusted image pair.
U.S. Patent No. 8,965,121 - "Image Color Matching and Equalization Devices and Related Methods"
- Technology Synopsis: The patent addresses color inconsistencies that can arise when capturing related images, for example, in panoramic stitching or stereo pairs, due to changes in lighting or differences in camera sensors (’121 Patent, col. 1:26-44). The claimed solution generates an array of color channel differences from overlapping portions of the images, uses a quantization technique to create a sparse look-up table of these differences, interpolates values to fill the table, and then applies the resulting color corrections to a target image ('121 Patent, Abstract).
- Asserted Claims: At least Claim 1 is asserted (Compl. ¶26).
- Accused Features: The complaint accuses the "noise reduction" and "color enhancement" features of Topaz Photo AI (Compl. ¶26-27).
III. The Accused Instrumentality
Product Identification
- The accused instrumentalities are Defendant's software products: Topaz Photo AI, Topaz Video AI, and Topaz Gigapixel AI (Compl. ¶1).
Functionality and Market Context
- The complaint describes the products as "AI-powered software for photo and video enhancement" (Compl. ¶7). Specific functionalities alleged to be relevant include:
- Topaz Video AI: Upscaling, frame rate conversion, stabilization, and noise reduction (Compl. ¶7).
- Topaz Photo AI: Noise reduction, sharpening, upscaling, face recovery, color enhancement, and batch processing (Compl. ¶7, ¶18, ¶26).
- Topaz Gigapixel AI: Image upscaling (Compl. ¶7).
- The complaint alleges that Defendant serves over one million customers globally, including major entities such as Apple Inc. and NASA (Compl. ¶5).
IV. Analysis of Infringement Allegations
No probative visual evidence provided in complaint.
’971 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| receiving a two-dimensional video sequence including a plurality of frames | Topaz Video AI processes input video files, which are described as two-dimensional sequences with multiple frames. | ¶9 | col. 2:9-11 |
| selecting a target frame from among the plurality of frames | The software selects specific frames for enhancement, which the complaint alleges is implied by features like "frame rate conversion." | ¶10 | col. 2:11-12 |
| selecting a first subset of frames...representative of a large stereo displacement | The software analyzes a subset of frames with a larger stereo baseline to extract depth data for static objects, which the complaint alleges is indicated by its "stabilization" and "enhancement" features. | ¶11 | col. 2:12-16 |
| extracting depth data of static objects utilizing the first subset of frames | The software is alleged to infer depth of static elements as part of its "AI-driven enhancement" and stabilization processes. | ¶12, ¶13 | col. 2:18-20 |
| selecting a second subset of frames...representative of a stereo displacement substantially different | The software uses a smaller subset of frames, such as adjacent frames, for processing moving objects. | ¶14 | col. 2:20-24 |
| utilizing the second subset of frames to calculate depth of moving objects | The software processes motion data to generate interpolated frames, which the complaint alleges implies depth calculations for moving objects. | ¶15 | col. 2:20-24 |
| generating a three-dimensional video frame based on the depth data | The software outputs enhanced video frames with improved depth perception, which are marketed as having "enhanced depth." | ¶16 | col. 2:25-27 |
Identified Points of Contention
- Technical Questions: A primary question is whether the accused product's general-purpose features (e.g., "stabilization," "frame rate conversion") actually perform the specific technical steps recited in the claim. For instance, the complaint alleges that the need to calculate depth is "implied" by or "required for" these features (Compl. ¶10, ¶15). The case may turn on evidence demonstrating that Topaz Video AI's underlying algorithms, in fact, select distinct large and small subsets of frames to separately calculate depth for static and moving objects, as opposed to using a different technical method to achieve stabilization or frame interpolation.
- Scope Questions: The analysis may raise the question of whether an "enhanced" 2D video frame with improved "depth perception" (Compl. ¶16) meets the claim limitation of a "three-dimensional video frame."
’580 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| receiving a plurality of captured images | Topaz Photo AI processes multiple images or frames, for instance, in its "batch processing" mode. | ¶18 | col. 5:53-56 |
| determining matching image features among the plurality of captured images | The software identifies common features like edges and textures across images, which the complaint alleges is implied by its "sharpening" and "face recovery" features. | ¶19 | col. 7:55-58 |
| ...applying the transformation...to create a plurality of transformed images | The software applies transformations to align features for consistent enhancement as part of its "AI-driven upscaling process." | ¶20, ¶21 | col. 13:45-50 |
| defining a transformed image pair | The software creates pairs of enhanced images for stereoscopic effect, which the complaint alleges is "implied by its ability to enhance depth perception." | ¶22 | col. 2:29-34 |
| calculating parallax values of the matched image features | The software assesses disparities between aligned images to adjust depth as part of its "sharpening" and "upscaling" algorithms. | ¶23 | col. 15:35-40 |
| ...altering a stereoscopic characteristic...to create an adjusted image pair | When disparities are determined to be excessive, the software adjusts pixel attributes to optimize depth, which is described as "AI-driven quality maximization." | ¶24 | col. 9:10-15 |
| defining a three-dimensional image based on the adjusted image pair | The software outputs enhanced images described as having "stereoscopic depth" and "professional-quality results." | ¶25 | col. 9:16-19 |
Identified Points of Contention
- Technical Questions: The central technical question appears to be whether the AI-based "upscaling" and "sharpening" algorithms in the accused products operate by performing the specific sequence of claim steps. The complaint frequently states that the claimed functionality is "implied by" the accused features (Compl. ¶19, ¶22). A dispute may arise over whether the accused AI processes actually create and manipulate a "transformed image pair" and calculate "parallax values," or if they use a different, single-image-focused neural network approach to enhance perceived depth and sharpness.
- Scope Questions: A likely point of contention will be whether "assessing disparities between aligned images" (Compl. ¶23) in an AI upscaling algorithm is the same as "calculating parallax values" in the context of stereoscopic imaging.
V. Key Claim Terms for Construction
For the ’388 Patent
The Term
- "extracting depth data"
Context and Importance
- This term is critical because the complaint links it to general features like "stabilization" and "AI-driven enhancement" (Compl. ¶11, ¶13). The viability of the infringement theory may depend on whether the processes underlying these features can be characterized as "extracting depth data." Practitioners may focus on this term because Defendant could argue its AI models enhance images without explicitly extracting depth maps in the manner disclosed in the patent.
Intrinsic Evidence for Interpretation
- Evidence for a Broader Interpretation: The specification does not appear to provide a single, explicit definition, which may support an argument that the term should be given its plain and ordinary meaning, potentially encompassing any calculation that yields information about object distance.
- Evidence for a Narrower Interpretation: The detailed description and figures describe depth data extraction in the specific context of forming a "stereoscopic pair of frames" from which disparities are measured ('388 Patent, col. 2:18-20, Fig. 6). This could support a narrower construction requiring the generation of a stereoscopic pair as a prerequisite for extraction.
For the ’580 Patent
The Term
- "altering a stereoscopic characteristic"
Context and Importance
- The infringement allegation for this element relies on the accused software's "AI-driven quality maximization" feature, which "adjusts pixel attributes to optimize depth" (Compl. ¶24). The dispute will likely center on whether this AI-based adjustment constitutes an alteration of a "stereoscopic characteristic." Practitioners may focus on this term to determine if general image quality optimization falls within the scope of a claim term directed at manipulating properties of a stereoscopic image pair.
Intrinsic Evidence for Interpretation
- Evidence for a Broader Interpretation: Claim 1 links the alteration to when "parallax values do not meet a predetermined criteria," suggesting any alteration performed to correct parallax issues could be covered. The specification states this can be done by adjusting "an attribute of at least one pixel" ('580 Patent, col. 9:12-13), which is broad language.
- Evidence for a Narrower Interpretation: The specification provides specific examples of altering stereoscopic characteristics, such as "screen plane alteration" and "object parallax adjustment" ('580 Patent, col. 10:33-42, Figs. 3D, 11, 12). These specific embodiments could be used to argue that the term is limited to direct manipulation of parallax in a registered image pair, rather than general AI-based pixel adjustments.
VI. Other Allegations
Indirect Infringement
- The complaint alleges active inducement, stating that Defendant provides "tutorials, user guides, and other implementation resources on its website" that instruct customers on using the infringing features (Compl. ¶33, ¶47). It alleges contributory infringement on the basis that Defendant's software contains modules that are material components "exclusively designed and marketed for the infringing functionalities, with no substantial non-infringing use" (Compl. ¶37, ¶47).
Willful Infringement
- Willfulness is alleged based on both pre-suit and post-suit knowledge. The complaint alleges pre-suit knowledge "on information and belief" from Defendant's general monitoring of the industry (Compl. ¶42). More specifically, it alleges Defendant had actual knowledge of the asserted patents at least as of the date it received Plaintiff's pre-suit demand letter and continued its allegedly infringing conduct thereafter (Compl. ¶2, ¶33, ¶53).
VII. Analyst’s Conclusion: Key Questions for the Case
- A central evidentiary question will be one of technical operation: does the accused software's use of artificial intelligence for image enhancement (e.g., stabilization, upscaling) perform the specific, multi-step methods recited in the claims—such as creating and processing stereoscopic pairs, calculating parallax from distinct baselines, and altering stereoscopic characteristics—or do the accused AI algorithms achieve their results through a fundamentally different technical process?
- A core issue will be one of definitional scope: can claim terms rooted in the language of conventional stereoscopic image processing, such as an "adjusted image pair" ('580 Patent) or extracting depth data from a "stereo displacement" ('388 Patent), be construed to encompass the implicit calculations and outputs of modern AI-based enhancement models that operate differently?
- The case may also test the sufficiency of infringement pleadings that rely heavily on inference, where claimed technical steps are alleged to be "implied by" or "indicated by" the high-level functionality of complex "black box" AI systems, pending further discovery into their internal workings.