PTAB

IPR2022-00490

Meta Platforms Inc v. EyesMatch Ltd

Log In
Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Method for Image Transformation, Augmented Reality, and Teleperence
  • Brief Description: The ’110 patent relates to imaging and display systems that digitally manipulate a camera-captured image of a user to simulate the appearance of a conventional reflective mirror. The patent discloses further enhancements, such as manipulating objects or the background in the captured images.

3. Grounds for Unpatentability

Ground 1: Obviousness over Core Mirror Simulation References - Claims 1, 5-6, and 10-11 are obvious over Rosenberg in view of Francois, Rosenthal, and Jeon.

  • Prior Art Relied Upon: Rosenberg (Application # 2007/0040033), Francois (a 2003 conference proceeding), Rosenthal (Application # 2007/0242066), and Jeon (Patent 6,677,980).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that the combination of these references teaches every limitation of the challenged claims. Rosenberg, the primary reference, discloses a "digital mirror system" with a camera, display, and processor that captures real-time images and transforms them (e.g., left-right inversion, scaling) to mimic a mirror. To meet the "adaptive transformation mapping" and "different point of view" limitations of claim 1, Petitioner relied on Francois, which teaches specific mathematical equations to transform an image from a camera's viewpoint to a user's viewpoint, creating a more convincing mirror simulation by accounting for the user's position. For the limitation of detecting a user in the stream of images, Petitioner argued Rosenberg’s motion/proximity sensing was sufficient, but also cited Rosenthal’s software-based facial feature tracking as an obvious alternative. Finally, to fully address the requirement that the modified image appear from a "different angle," Petitioner cited Jeon, which teaches "gaze correction" techniques that virtually rotate a captured image to correct the user's gaze, thereby modifying the image to appear as if captured from a different angle.
    • Motivation to Combine: A POSITA starting with Rosenberg's basic digital mirror would combine it with Francois to implement a more sophisticated and convincing mirror effect using Francois's straightforward and efficient transformation equations. A POSITA would have been motivated to incorporate Rosenthal's facial detection techniques to enable compelling features like overlaying virtual objects (e.g., virtual glasses), a feature also suggested by Rosenberg (virtual hairstyles). Furthermore, a POSITA would combine Jeon to solve the well-known problem of poor eye contact in systems where the camera is not at eye-level, a limitation inherent in the Rosenberg and Francois systems.
    • Expectation of Success: A POSITA would have a reasonable expectation of success because combining the references involved applying known image processing techniques to solve predictable problems and add desirable, known features.

Ground 2: Obviousness over Core Combination plus Telepresence Art - Claim 12 is obvious over Rosenberg, Francois, Rosenthal, and Jeon, in further view of Agamanolis.

  • Prior Art Relied Upon: The prior art from Ground 1, and Agamanolis (a 1997 conference proceeding).
  • Core Argument for this Ground:
    • Prior Art Mapping: Claim 12 adds the limitation of displaying the modified image on a remote screen for "video telepresence." Petitioner asserted that Agamanolis teaches this limitation through its "magic mirror" system, which creates a telepresence environment by segmenting multiple remote users from their backgrounds and combining their images into a single shared virtual space. This composite image is then distributed for display on the remote screens of each participant, allowing them to see themselves and others as if they were in the same room.
    • Motivation to Combine: A POSITA would combine the mirror simulation system from Ground 1 with Agamanolis's telepresence features to extend the application from a single-user experience to a more realistic, multi-user, networked communication environment. Agamanolis expressly states its system enhances communication, providing a clear motivation to incorporate its features into the networked mirror system taught by Rosenberg.
    • Expectation of Success: The techniques in Agamanolis were well-known years before the ’110 patent's priority date, ensuring a POSITA would have a high expectation of success in implementing them.

Ground 3: Obviousness over Core Combination plus Background Segmentation Art - Claims 16-17 are obvious over Rosenberg, Francois, Rosenthal, and Jeon, in further view of Baron and Agamanolis.

  • Prior Art Relied Upon: The prior art from Ground 1, Baron (Patent 8,823,739), and Agamanolis.
  • Core Argument for this Ground:
    • Prior Art Mapping: Claims 16 and 17 add limitations for generating "background segmentation" and replacing the background with a "virtual background or background that was captured in a different setting." Petitioner argued that both Baron and Agamanolis teach these features. Baron explicitly discloses methods for isolating a user's face or torso from their background in a video stream and merging the user's image with a desired virtual background (e.g., a corporate logo, a webpage, or a video feed). Agamanolis similarly teaches extracting a participant from their surroundings using a segmentation algorithm to place them in a shared virtual background.
    • Motivation to Combine: A POSITA would be motivated to add virtual background capabilities to the Ground 1 mirror system to obtain the well-understood benefits of this feature. As explained by Baron, virtual backgrounds enhance the user experience by replacing potentially distracting or unprofessional real-world backgrounds with more attractive or relevant imagery, a highly desirable feature for any video communication system.
    • Expectation of Success: Virtual background technology was conventional at the time, and a POSITA would have readily implemented the processes described in Baron and Agamanolis using standard programming techniques.

4. Arguments Regarding Discretionary Denial

  • Petitioner argued against discretionary denial under Fintiv, stating that the parallel district court litigation was in its early stages, with a trial date more than 20 months away. Petitioner also asserted its intent to move to stay the litigation and stipulated it would not pursue any instituted ground in court.
  • Petitioner further argued against denial under §325(d), contending that the petition raised new, non-cumulative arguments. The key references of Francois, Jeon, Agamanolis, and Baron were not before the Examiner during prosecution. While Rosenberg was cited in an Information Disclosure Statement (IDS), it was never applied in any rejection, and the Examiner allowed the claims in the first office action before the IDS containing Rosenberg was even submitted.

5. Relief Requested

  • Petitioner requests institution of an IPR and cancellation of claims 1, 5-6, 10-12, and 16-17 of the ’110 patent as unpatentable.