PTAB

IPR2025-01177

Google LLC v. SoundClear Technologies LLC

Log In
Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Output-Content Control Device, Method, and Program
  • Brief Description: The ’675 patent discloses an "output-content control device" that generates responses to spoken user inputs. The device classifies a user's voice based on distance (e.g., a "first voice" for close proximity/whisper or a "second voice" for normal speech) and modifies the content of the generated response to enhance user privacy.

3. Grounds for Unpatentability

Ground 1: Claims 1-7 are obvious over Raitio in view of Fahlman.

  • Prior Art Relied Upon: Raitio (Application # 2017/0358301) and Fahlman (Patent 5,960,080).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Raitio, which discloses a digital assistant, teaches nearly all limitations of the independent claims. Raitio’s system was asserted to acquire a user's voice, analyze it to infer intent, and generate a response. Critically, Raitio’s assistant can classify the input as whispered or non-whispered (a first or second voice) using various acoustic characteristics and can implement a "far-field acoustic model" that uses a distance-based threshold, allegedly teaching the claimed proximity sensing and distance calculation. Raitio’s system also generates different outputs based on this classification (e.g., a whispered response for a whispered input) to protect user privacy. Petitioner contended the only element not expressly taught by Raitio is replacing a specific word in the output sentence with another word when the input is classified as the "first voice." To supply this feature, Petitioner asserted Fahlman teaches a method for sanitizing messages containing sensitive information by replacing specific terms with "standard tokens" to preserve confidentiality.
    • Motivation to Combine: Petitioner argued a person of ordinary skill in the art (POSITA) would combine Raitio and Fahlman to enhance the privacy features already central to Raitio's disclosure. Since Raitio already modifies its output (by whispering) to protect privacy, a POSITA would find it obvious to incorporate Fahlman’s word-replacement technique to further sanitize the response and prevent private information from being overheard, even in a whisper.
    • Expectation of Success: A POSITA would have a reasonable expectation of success because both Raitio and Fahlman describe software-based systems. Implementing Fahlman's word-replacement algorithm within Raitio's dialogue generation process would involve the routine integration of known software modules.

Ground 2: Claims 1-7 are obvious over Ocampo in view of Fahlman.

  • Prior Art Relied Upon: Ocampo (Application # 2018/0122361) and Fahlman (Patent 5,960,080).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner asserted that Ocampo discloses a user device, such as a personal digital assistant, that performs the core functions of the claimed invention. Ocampo's device was said to acquire a user's voice via a microphone, analyze the command to determine user intent using an "application determiner," and classify the user's voice based on various features. This classification includes determining if a user is whispering and using a "proximity classifier" to determine the user's distance from the device. Based on this classification, Ocampo generates a corresponding audio output, such as a low-volume or whispered response, to avoid user embarrassment or inconvenience to others. Similar to the first ground, Petitioner argued that Ocampo teaches all limitations except for replacing a word in the output with another word. Fahlman was again cited for its teaching of replacing sensitive terms with standard tokens to create a "sanitized message."
    • Motivation to Combine: The motivation presented was parallel to Ground 1. Ocampo explicitly recognized a user's desire for privacy and discretion (e.g., whispering to avoid embarrassment). Petitioner argued a POSITA would be motivated to integrate Fahlman's sanitization method into Ocampo's system to further this goal. By replacing potentially sensitive words in the response, the system would better protect user privacy beyond simply lowering the volume.
    • Expectation of Success: Petitioner claimed a POSITA would expect success in this combination. Ocampo's system architecture, which includes various software modules like an "audio template selector," could readily incorporate Fahlman’s software-based algorithm for word replacement with only routine skill.

4. Relief Requested

  • Petitioner requests the institution of an inter partes review and the cancellation of claims 1-7 of the ’675 patent as unpatentable.