DCT
5:25-cv-02964
Google LLC v. EyesMatch Ltd
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: Google LLC (Delaware)
- Defendant: EyesMatch Ltd. (British Virgin Islands)
- Plaintiff’s Counsel: Arnold & Porter Kaye Scholer LLP
- Case Identification: 3:25-cv-02964, N.D. Cal., 03/31/2025
- Venue Allegations: Plaintiff Google asserts that venue is proper in the Northern District of California because a substantial part of the events giving rise to the claim occurred in the district and Defendant EyesMatch is subject to personal jurisdiction there, based on alleged business contacts, licensing activities to a Palo Alto-based company, and enforcement activities directed at Google within the district.
- Core Dispute: Plaintiff Google seeks a declaratory judgment that its Google Meet application does not infringe two patents owned by Defendant EyesMatch related to image processing for creating a virtual mirror experience.
- Technical Context: The technology at issue concerns the digital manipulation of a video stream from a camera to simulate the visual properties of a physical mirror or to alter the user's perceived point of view, a key user-experience feature in modern video conferencing and augmented reality applications.
- Key Procedural History: This declaratory judgment action follows a patent infringement lawsuit filed by EyesMatch against Samsung in the Eastern District of Texas, which alleged that the Google Meet application preinstalled on Samsung devices infringes the patents-in-suit. EyesMatch subsequently served a subpoena on Google in California related to the Texas litigation, prompting Google to file this action seeking a declaration of non-infringement. EyesMatch previously litigated and settled a case involving the same patents against Facebook, Inc.
Case Timeline
| Date | Event |
|---|---|
| 2005-03-01 | Earliest Priority Date for '109 and '110 Patents |
| 2013-03-15 | '109 Patent Application Filing Date |
| 2014-04-15 | '110 Patent Application Filing Date |
| 2015-03-17 | '109 Patent Issue Date |
| 2015-03-17 | '110 Patent Issue Date |
| 2021-01-28 | EyesMatch files infringement suit against Facebook et al. |
| 2023-08-08 | EyesMatch files infringement suit against Samsung et al. |
| 2024-04-24 | EyesMatch serves subpoena on Google in California |
| 2025-03-31 | Google files Complaint for Declaratory Judgment |
II. Technology and Patent(s)-in-Suit Analysis
U.S. Patent No. 8,982,109 - "Devices, Systems and Methods of Capturing and Displaying Appearances," issued March 17, 2015
The Invention Explained
- Problem Addressed: The patent's background section describes that an image from a standard camera displayed on a screen is not a convincing substitute for a conventional mirror because the user's reflection does not behave as it would in a real mirror, particularly with respect to how the field of view and image size change as the user moves closer or farther away (’109 Patent, col. 2:2-27).
- The Patented Solution: The invention describes a system that captures a video feed of a user and applies a series of digital manipulations to create a "convincing mirror experience" on a display (’109 Patent, col. 14:14-19). This is achieved by flipping the image horizontally to mimic left-right reversal and, crucially, by performing transformations like resizing the image to compensate for the user's changing distance from the camera, thereby making the on-screen reflection behave more like a physical mirror (’109 Patent, col. 14:31-40).
- Technical Importance: The described method addresses a persistent user-experience challenge in video conferencing and virtual try-on systems by making the digital self-view feel more intuitive and natural.
Key Claims at a Glance
- The complaint identifies independent claim 1 as the allegedly infringed claim (Compl. ¶21).
- The essential elements of Claim 1 include:
- obtaining a digital image from a camera;
- flipping the image about a vertical axis;
- applying a transformation mapping to mimic a mirror reflection;
- resizing the image to reduce variations from changes in the object's distance;
- displaying the processed image;
- performing these steps on a live video feed;
- determining the distance to the user; and
- varying the rate at which the steps are performed on the images according to the determined distance.
- The complaint does not explicitly reserve the right to assert dependent claims but seeks a judgment of non-infringement for "any claim of the '109 patent" (Compl. ¶25-26).
U.S. Patent No. 8,982,110 - "Method for Image Transformation, Augmented Reality, and Teleperence," issued March 17, 2015
The Invention Explained
- Problem Addressed: The patent addresses the technical challenge of making video conferencing more natural, noting that when a camera is placed above a screen, the user appears to be looking down rather than making eye contact ('110 Patent, col. 17:5-14). It also addresses creating more realistic virtual mirror and augmented reality experiences.
- The Patented Solution: The invention is a method that applies an "adaptive transformation mapping" to a video stream to generate modified images that "appear to be captured from a different point of view of the camera's actual point of view" ('110 Patent, Abstract). By creating a virtual camera perspective, the system can correct for gaze angle in telepresence or create more realistic augmented reality effects ('110 Patent, col. 1:1-2; col. 23:1-8).
- Technical Importance: This technology aims to improve the realism and engagement of digital interactions by manipulating the user's perceived perspective, a foundational concept for advanced telepresence and augmented reality.
Key Claims at a Glance
- The complaint identifies independent claim 1 as the allegedly infringed claim (Compl. ¶29).
- The essential elements of Claim 1 include:
- generating a stream of images of a user;
- detecting the presence of the user;
- applying an adaptive transformation mapping to generate modified images that appear to be from a different point of view than the camera's actual point of view;
- displaying the modified images;
- wherein the different point of view corresponds to an image taken from a different angle and a different distance than the actual images; and
- wherein during mirror tracking mode, the processor uses the angle of the captured images and applies parametric input for a mapping transformation.
- The complaint seeks a judgment of non-infringement for "any claims of the '110 patent" (Compl. ¶32, 34).
III. The Accused Instrumentality
Product Identification
- The "Auto Framing feature" of the "preinstalled Google Meet application (formerly Google Duo)" on Samsung smartphones and other devices (Compl. ¶3, 19, 29).
Functionality and Market Context
- The complaint, filed by Google, does not describe how the accused "Auto Framing" feature functions. Instead, it defines the feature by what it allegedly does not do in order to rebut infringement. The complaint contends the feature does not measure the distance of a user from the camera and does not vary any processing rate based on distance (Compl. ¶23). It further contends the feature does not generate an image from a different angle than the actual camera's angle (Compl. ¶31). The Google Meet application is a widely used video conferencing service, and its inclusion on devices from major manufacturers like Samsung gives this dispute significant commercial relevance (Compl. ¶3, 5). No probative visual evidence provided in complaint.
IV. Analysis of Infringement Allegations
The complaint seeks a declaratory judgment of non-infringement. The tables below summarize Google's asserted non-infringing functionality against the key limitations of the asserted claims.
'109 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Google's Contended Non-Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| determining distance to a user appearing in the live video feed | "The accused auto-framing functionality does not measure the distance of the user from the camera." | ¶23 | col. 29:63-65 |
| varying rate at which the non-ordered steps are performed on the series of images according to the distance | "Google does not perform 'varying rate at which the non-ordered steps are performed on the series of images according to the distance.' For example, there is no varying of a rate 'according to the distance' of a user." "The Google Meet application does not perform any of these steps at a varying rate according to distance." | ¶23, 18 | col. 30:1-4 |
'110 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Google's Contended Non-Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| operating the processor to apply adaptive transformation mapping ... to generate modified images that appear to be captured from a different point of view of the camera's actual point of view | "Specifically, the accused auto-framing functionality does not generate a modified image at a different angle than the actual image captured by the camera." | ¶31 | col. 36:58-63 |
| wherein the different point of view corresponds to an image taken from a different angle and a different distance than the actual images captured by the camera | "Instead, the modified image has the same angle as the actual image captured by the camera. The modified image never has a different point of view based on a different angle." | ¶31 | col. 36:64-67 |
- Identified Points of Contention:
- Technical Questions: The primary dispute for both patents appears to be factual and technical. For the ’109 patent, the question is whether the Google Meet auto-framing feature in fact "determin[es] distance" and "vary[es] rate... according to the distance." For the ’110 patent, the question is whether the digital transformations used for auto-framing (e.g., cropping, digital zoom) result in an image from a "different angle" than the physical camera.
- Scope Questions: The dispute over the ’110 patent raises a significant claim scope question: Does the term "different angle," in the context of creating a "different point of view," read on the output of digital image processing like panning and zooming, or is it limited to a geometric transformation that simulates a physical repositioning of the camera?
V. Key Claim Terms for Construction
The Term: "varying rate at which the non-ordered steps are performed ... according to the distance" (’109 Patent, Claim 1)
- Context and Importance: This limitation is central to Google's non-infringement argument for the ’109 patent. Practitioners may focus on this term because its construction will determine whether any correlation between processing adjustments and user distance is sufficient, or if a specific, measured dependency is required.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The patent's objective is to "create a convincing mirror appearance" (’109 Patent, col. 3:59-60). A party could argue that any processing adjustments that have the effect of compensating for distance, even if not directly calculated from a distance measurement, fall within the scope.
- Evidence for a Narrower Interpretation: The claim explicitly requires "determining distance" as a distinct step. The specification links this to controlling camera zoom and focus in real-time "based on user tracking" (’109 Patent, col. 14:38-40), which may support an interpretation requiring a direct functional link between a measured distance and a variable processing rate.
The Term: "different point of view... from a different angle" (’110 Patent, Claim 1)
- Context and Importance: This is the dispositive term for the ’110 patent dispute, as Google contends its auto-framing does not change the "angle" of the image. The case may turn on whether digital cropping and scaling falls within the patent's definition of creating a new viewing "angle."
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The patent abstract states the goal is to generate images that "appear to be captured from a different point of view." One could argue that a digital zoom that centers a user's face creates the appearance of a different point of view, satisfying the term's purpose.
- Evidence for a Narrower Interpretation: The specification discusses correcting for geometric "projection distortion" caused by a tilted camera (’110 Patent, col. 15:29-34). It also explicitly mentions applying "tilt transformation" and "elevation transformation" (’110 Patent, col. 23:45-54). This language suggests the term "different angle" refers to a geometric, perspective-altering transformation, not merely cropping or scaling an existing image.
VI. Other Allegations
- Indirect Infringement: The complaint asserts non-infringement of any indirect infringement claims by arguing that because there is no underlying direct infringement by Google Meet users, Google cannot be liable for inducement or contributory infringement (Compl. ¶24, 32). It further denies the requisite specific intent for inducement and knowledge for contributory infringement (Compl. ¶24, 32).
- Willful Infringement: While this is a declaratory judgment action and no willfulness is alleged by the patentee herein, the complaint acknowledges facts that could form the basis of a future willfulness claim. Google’s awareness of the patents and EyesMatch's infringement position is established by the prior lawsuit against Samsung and the subpoena served on Google (Compl. ¶2-3).
VII. Analyst’s Conclusion: Key Questions for the Case
- A central issue will be one of technical evidence: Does the accused Google Meet auto-framing feature, as it actually operates, incorporate a mechanism for determining a user’s distance from the camera and varying a processing rate based on that distance, as required by the '109 patent? The outcome may depend entirely on the evidence produced regarding the software's internal logic.
- The second key issue is one of claim construction: Can the term "different point of view... from a different angle" in the '110 patent be interpreted to encompass the effects of digital image processing such as cropping and scaling, or does the patent’s specification limit the term to a more fundamental geometric transformation of the image's perspective?