Google LLC v. Cellular South Inc

1. Case Identification

• Case #: IPR2025-00875
• Patent #: 9,940,972
• Filed: May 10, 2025
• Petitioner(s): Google LLC
• Patent Owner(s): Cellular South, Inc.
• Challenged Claims: 1-20

2. Patent Overview

• Title: Video to Data
• Brief Description: The ’972 patent discloses methods and systems for generating data from video content. The technology involves processing a video’s audio and image components in parallel to extract information, generate metadata, and create additional content, such as context-aware advertisements.

3. Grounds for Unpatentability

Ground 1: Claims 1-20 are obvious over Fontana in view of Lau.

• Prior Art Relied Upon: Fontana (Application # 2012/0078712) and Lau (Application # 2007/0112630).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Fontana taught the core elements of independent claims 1 and 17, including a method and system for processing multimedia content. Fontana disclosed separating audio and video streams, performing parallel processing on a distributed computing system, extracting objects from image files (e.g., thumbnails), and converting audio to text. However, Fontana only broadly mentioned linking advertisements to content. Petitioner asserted that Lau supplied the missing element by teaching a detailed method for matching advertisements to video content. Lau described cross-referencing keywords and concepts from a video’s text and image data to determine relevant topics and then selecting and placing ads based on those topics. The combination of Fontana’s content processing framework with Lau’s advertisement matching and placement system allegedly rendered all limitations of the independent claims obvious. Dependent claims were argued to be obvious as they recited known techniques like natural language processing, facial recognition, and optical character recognition (OCR), all of which were disclosed or suggested by Fontana and Lau.
  • Motivation to Combine: A Person of Ordinary Skill in the Art (POSITA) would combine Fontana and Lau to implement a complete system for content-aware advertising. Fontana expressly disclosed that advertisement data could be used "to link one or more advertisements with multimedia content during playback" but did not detail the matching process. Lau provided a specific, keyword-based search and matching method for this exact purpose. A POSITA would therefore look to a reference like Lau to supply the advertisement-matching functionality suggested but not detailed by Fontana, thereby creating a more complete and commercially valuable system.
  • Expectation of Success: A POSITA would have a reasonable expectation of success in combining the references. Both Fontana and Lau operated in the same technical field of multimedia content processing and advertising. Integrating Lau’s topic-matching algorithm into Fontana’s modular, distributed processing architecture was presented as a predictable and straightforward implementation of known technologies to achieve a known goal.

Ground 2: Claims 8-9 and 20 are obvious over Fontana in view of Lau and Arakawa.

• Prior Art Relied Upon: Fontana (Application # 2012/0078712), Lau (’630 application), and Arakawa (Application # 2012/0239401).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground built upon the combination of Fontana and Lau from Ground 1 and added Arakawa to address the specific limitations of claims 8, 9, and 20. These claims require segmenting audio and/or image files before parallel processing, with claims 9 and 20 specifying segmentation at "spectrum thresholds." Petitioner argued that Arakawa taught an advanced method for improving speech recognition in noisy environments by first segmenting audio into voice and non-voice sections based on a threshold value. This pre-processing segmentation, which Arakawa taught occurs before the main recognition process, directly corresponded to the claimed step of "segmenting the audio files before processing the audio files in parallel." Arakawa’s use of a threshold to distinguish voice from non-voice frames was asserted to be a form of segmentation at a spectrum threshold.
  • Motivation to Combine: A POSITA implementing the Fontana/Lau system would be motivated to incorporate Arakawa’s teachings to improve the system’s performance. Fontana explicitly taught that "a plurality of different speech to text algorithms can be applied." A POSITA would seek to use more effective algorithms, and Arakawa provided a method to enhance speech recognition accuracy by filtering out background noise through pre-processing segmentation. This would lead to better text transcripts, more accurate topic identification, and more relevant ad placements. Furthermore, Fontana suggested segmenting content based on "sound changes," motivating a POSITA to use Arakawa's threshold-based method to segment both audio and corresponding image files.
  • Expectation of Success: Integrating Arakawa’s audio segmentation algorithm into Fontana’s system would be predictable. Arakawa’s teachings represented a known technique for improving a standard component (speech-to-text conversion) within the broader multimedia processing framework described by Fontana.

4. Arguments Regarding Discretionary Denial

• Petitioner argued against discretionary denial under 35 U.S.C. §314(a) and §325(d).
• Fintiv Factors (§314(a)): Petitioner contended that the co-pending litigation (Cellular South, Inc. v. Google LLC, N.D. Cal.) was in a very early stage, with no substantive discovery, claim construction, or trial date set. Petitioner also stipulated that it would not pursue the same invalidity grounds in the district court if the inter partes review (IPR) is instituted.
• Advanced Bionics Factors (§325(d)): Petitioner argued that the primary prior art references, Fontana and Arakawa, were never presented to or considered by the Examiner during the original prosecution of the ’972 patent. While Lau was mentioned in the file history, it was only cited in passing by the Examiner for a different purpose and was not applied in any rejection or considered in combination with Fontana.

5. Relief Requested

• Petitioner requested institution of an IPR and cancellation of claims 1-20 of the ’972 patent as unpatentable under 35 U.S.C. §103.