US 8355484

A concise summary of US Patent 8,355,484, including its involvement in recent intellectual property litigation, is provided below.

Summary of US Patent 8,355,484

Title: Methods and apparatus for masking latency in text-to-speech systems

Assignee: The current assignee is listed as Cerence Operating Co.

Inventors: Ellen Marie Eide, Wael Mohamed Hamza

Filing Date: January 8, 2007

Issue Date: January 15, 2013

Abstract: A technique for masking latency in an automatic dialog system is provided. A communication is received from a user at the automatic dialog system. The communication is processed in the automatic dialog system to provide a response. At least one transitional message is provided to the user from the automatic dialog system while processing the communication. A response is provided to the user from the automatic dialog system in accordance with the received communication from the user.

Plain-Language Overview of Independent Claims

This patent has three independent claims: Claim 1 (an automatic dialog system), Claim 3 (an apparatus for producing speech output), and Claim 6 (a method of producing speech output).

• Claim 1: An Automatic Dialog System
  In simple terms, this claim describes a voice-activated system that uses "filler" sounds or words to hide processing delays. When you speak to the system, a "filler generator" immediately picks a pre-recorded sound (like "um" or a cough) or a short phrase (like "let's see..."). A speech synthesizer then plays this "transitional message" to you. This continues until the system has figured out the real answer to your query. Once the actual response is ready, the system stops playing the filler messages and gives you the answer. The core idea is to make the system's thinking time feel more natural, like a person pausing in conversation, rather than having awkward silences.

• Claim 3: An Apparatus for Producing Speech Output
  This claim covers the hardware components that make the latency-masking system work. It describes an apparatus with a memory and at least one processor. The processor is programmed to receive a user's spoken communication, process it to come up with a response, and in the meantime, play one or more "transitional messages" to mask any delay. A key feature is that a "filler generator" randomly selects these transitional messages from a database. This apparatus uses a speech synthesis system to produce both the filler messages and the final, meaningful response.

• Claim 6: A Method of Producing Speech Output
  This claim focuses on the process or steps involved in masking the delay. The method involves a "filler generator" selecting multiple transitional messages after receiving a user's communication. Then, a speech synthesis system plays these filler messages to the user to cover up the time the system is busy processing the request. Once the actual response is ready, the system stops the filler messages and delivers the response. This claim emphasizes that both the filler sounds/phrases and the final answer are audibly generated by the speech synthesis system.

Litigation Update

As of May 2026, Cerence Operating Co. has initiated patent infringement litigation against Amazon.com Inc. In a complaint filed with the U.S. International Trade Commission (ITC) on May 5, 2026, and a parallel lawsuit in the Eastern District of Texas, Cerence alleges that Amazon's products, including the Echo line of smart speakers and the Alexa voice assistant, infringe on several of its patents, including US 8,355,484. The specific claim is that these Amazon products use Cerence's patented technology for masking latency in text-to-speech conversion. Cerence is seeking to block the importation of the accused Amazon devices and to obtain monetary damages. At present, there is no authoritative information available from the CAFC 2026 dockets regarding this specific patent.