Google LLC v. 138 East LCD Advancements Ltd

1. Case Identification

• Case #: IPR2024-00960
• Patent #: 7,486,807
• Filed: June 20, 2024
• Petitioner(s): Google LLC
• Patent Owner(s): 138 East LCD Advancements Limited
• Challenged Claims: 1-14

2. Patent Overview

• Title: Image Retrieving Device, Method for Adding Keywords, and Computer Program
• Brief Description: The ’807 patent discloses a system and method for classifying and retrieving digital images. The system detects an object within an inputted image, proposes keywords related to that object, and allows a user to confirm, add, or correct the keywords, which are then stored with the image in a database.

3. Grounds for Unpatentability

Ground 1: Claims 1-14 are Anticipated or Obvious over Stubler

• Prior Art Relied Upon: Stubler (Application # 2002/0188602).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued Stubler discloses or renders obvious all limitations of independent claims 1, 7, and 8. Stubler teaches a system for generating "captions or semantic labels" (keywords) for images by detecting "regional features" such as faces or skin (objects) and comparing them to a database of stored images. The system stores images with their associated keywords in a database, detects newly inputted images that are "unlabeled" or "not yet been annotated," and proposes keywords from similar stored images for the new image. Stubler further discloses an "interactive user verification" process where a list of proposed captions is presented on a display, allowing the user to "select and/or edit" the keywords. This mapping covers the core limitations of storing an image with a keyword, detecting a new image, acknowledging an object, proposing a keyword, and allowing a user to confirm, add, or correct it.
  • Motivation to Combine: For limitations Petitioner argued are not explicitly disclosed but would be obvious (e.g., an "image inputting detecting section"), a POSITA would implement a program for detecting a newly inputted image to automate the process and know when an image is ready for processing. This would further Stubler’s goal of "reducing the effort required of the user."
  • Expectation of Success: A POSITA would have reasonably expected success in implementing such features, as Stubler noted that the "software needed for implementation of the method is within the ordinary skill in such arts." The implementation would have been a straightforward application of known software operations to yield predictable results.

Ground 2: Claims 2-4 and 9-12 are Obvious over Stubler and Wang

• Prior Art Relied Upon: Stubler (Application # 2002/0188602) and Wang (Patent 6,038,333).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued this combination renders obvious the dependent claims related to specific human and face detection methods. While Stubler provides the base system for image classification and keyword management, Wang provides specific teachings on advanced face detection that satisfy the limitations of claims 2-4. Wang discloses a "face detection and feature extraction system" that uses a "neural network-based face detection scheme" to determine if an image contains a human face (claim 2). It further teaches retrieving "person-identifying data" such as a name or affiliation for a detected face and allowing a user to input such information if a match is not found, satisfying the "personal information inputting section" of claim 3. Stubler itself discloses, through incorporation by reference, using skin color detection when a human is detected (claim 4).
  • Motivation to Combine: A POSITA would combine Wang’s robust face detection technology with Stubler's system to improve the performance of a key feature envisioned by Stubler—detecting faces and assessing facial similarities. Wang’s neural network-based approach offered a more robust and faster solution than the general techniques mentioned in Stubler. A POSITA would also incorporate Wang's teaching on using "person-identifying data" to fulfill Stubler’s goal of applying "personally meaningful" labels to images, thereby improving image classification and retrieval.
  • Expectation of Success: A POSITA would have had a reasonable expectation of success, as the combination involves augmenting Stubler's software with a known type of functional module (Wang's face detection system). This integration of known software processes to perform their expected functions would have required only routine skill and would predictably result in an improved image classification system.

4. Arguments Regarding Discretionary Denial

• §314(a) / Fintiv: Petitioner argued that discretionary denial under Fintiv is not warranted because no parallel district court litigation is active. A prior related case was concluded with a final judgment, and no infringement or invalidity contentions concerning the ’807 patent were exchanged. Thus, the Fintiv factors, including the absence of a trial date and minimal party investment, strongly favor institution.
• §325(d): Petitioner argued that discretionary denial is inappropriate because the primary prior art references, Stubler and Wang, were not presented to or evaluated by the Examiner during the original prosecution. The petition therefore raises new arguments and art combinations that the Patent Office has not previously considered, warranting institution.

5. Relief Requested

• Petitioner requests the institution of an inter partes review and cancellation of claims 1-14 of the ’807 patent as unpatentable.