Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 1 of 23 PageID# 1
`
`IN THE UNITED STATES DISTRICT COURT
`FOR THE EASTERN DISTRICT OF VIRGINIA
`ALEXANDRIA DIVISION
`
`SOUNDCLEAR TECHNOLOGIES LLC,
`
`v.
`
`GOOGLE LLC,
`
`Plaintiff,
`
`Defendant.
`
`Case No. 1:24-cv-00729
`
`Jury Trial Demanded
`
`COMPLAINT FOR PATENT INFRINGEMENT
`
`Plaintiff SoundClear Technologies LLC (“SoundClear”) files this complaint against
`
`Google LLC (hereinafter collectively “Google” or “Defendant”) for infringement of United
`
`States Patent Nos. 9,031,259; 9,070,374; and 9,804,819 (the “Patents-in-Suit”) , attached here as
`
`Exhibits 1-3.
`
`NATURE OF THE ACTION
`
`1.
`
`This is an action for patent infringement arising under the patent laws of the
`
`United States, 35 U.S.C. §§1 et seq.
`
`THE PARTIES
`
`2.
`
`SoundClear is a limited liability company organized under laws of the
`
`Commonwealth of Virginia with its principal place of business at 1900 Reston Metro Plaza,
`
`Suite 600, Reston, VA 20190.
`
`1
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 2 of 23 PageID# 2
`
`3.
`
`On information and belief, defendant Google is a corporation organized and
`
`existing under the laws of the state of Delaware with a principal place of business at 1600
`
`Amphitheatre Parkway, Mountain View, California 94043.
`
`4.
`
`On information and belief, Google.com, Inc. may be served with process through
`
`its registered agent, Corporation Service Company, 251 Little Falls Drive, Wilmington,
`
`Delaware 19808, or anywhere it may be found.
`
`5.
`
`Google does business across the United States, including in the Commonwealth of
`
`Virginia and, more specifically, in the Eastern District of Virginia through at least its offices at
`
`1900 Reston Metro Plaza, Reston, Virginia 20190.
`
`6.
`
`On information and belief, Google has been authorized to transact business in the
`
`Commonwealth of Virginia and the Eastern District of Virginia since on or about October 31,
`
`2017, under Virginia Entity ID T0723124.
`
`7.
`
`On information and belief, Defendant sells and offers to sell products and services
`
`throughout Virginia, including in this judicial district, as well as throughout the United States,
`
`and introduces products and services that perform infringing processes into the stream of
`
`commerce knowing that they would be used, offered for sale, or sold in this judicial district and
`
`elsewhere in the United States.
`
`8.
`
`On information and belief, Google has made, used, offered to sell, offered to sell
`
`access to, sold, and/or sold access to products and services, including the following specifically
`
`accused products and services: (1) Google Home, Google Nest Mini (1st Gen), Google Home
`
`Mini (1st Gen), Google Home Max, Google Nest Audio, Google Nest Hub, Google Nest Hub
`
`Max, and Google Nest Hub (2nd Gen.), and Google Assistant (collectively “Google Home
`
`Products”); (2) current or legacy products or services, which use, or have used, one or more of
`
`
`
`2
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 3 of 23 PageID# 3
`
`the foregoing products and services as a component product or component service; (3)
`
`combinations of products and/or services comprising, in whole or in part, two or more of the
`
`foregoing products and services; and (4) all other current or legacy products and services
`
`imported, made, used, sold, or offered for sale by Google that operate, or have operated in a
`
`substantially similar manner as the above-listed products and services. (As used herein, one or
`
`more of the foregoing products and services are individually and collectively referred to as “the
`
`Google Products and Services”).
`
`9.
`
`On information and belief, Google, as well as the hardware and software
`
`components comprising the Google Products and Services and/or that enable the Google
`
`Products and Services to operate, including but not limited to servers, server software, webserver
`
`software, webserver hardware, email server hardware, email server software, website client
`
`software, mobile computing device client application software, networked communications
`
`hardware, network routers, network switches, network hubs, WIFI access point hardware, WIFI
`
`access point software, point-of-sale hardware, point-of-sale software, back-end hardware, back-
`
`end software, cloud-based software, cloud-based hardware, and other hardware and software
`
`computing systems and components infringes (literally and/or under the doctrine of equivalents)
`
`at least one claim of each of the Patents-in-Suit.
`
`JURISDICTION AND VENUE
`
`10.
`
`This civil action arises under the Patent Laws of the United States, 35 U.S.C. § 1
`
`et seq. Accordingly, this Court has subject matter jurisdiction under at least 28 U.S.C. §§ 1331
`
`and 1338(a).
`
`11.
`
`This Court has general and specific personal jurisdiction over Google because it
`
`regularly conducts and solicits business, or otherwise engages in other persistent courses of
`
`
`
`3
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 4 of 23 PageID# 4
`
`conduct in this judicial district, and/or derives substantial revenue from the use, sale, and
`
`distribution of goods and services, including but not limited to the accused Google Products and
`
`Services provided to individuals and businesses in the Eastern District of Virginia.
`
`12.
`
`Google has committed and continues to commit acts of infringement within this
`
`district and, thereby, giving rise to this action and establishing minimum contacts with this forum
`
`such that the exercise of jurisdiction over Google would not offend traditional notion of fair play
`
`and substantial justice. Google has employees, offices, and facilities in this District and has
`
`purposefully conducted and continues to purposefully conduct business in this District, as
`
`demonstrated by (a) Google’s maintenance of regular and established places of business in this
`
`District, including its office at 1900 Reston Metro Plaza, Reston, VA 20190 (see
`
`https://www.restonnow.com/2021/03/18/just-in-google-to-lease-more-space-at-reston-station/),
`
`(b) Google’s advertisement of 65 available job postings for its Reston office as of April 2023
`
`(see https://careers.google.com/locations/reston/), and (c) Google’s economic impact report
`
`stating that Google also provided “more than 475,000 Virginia businesses” with “direct
`
`connections to their customers” in 2021 including by, inter alia, providing directions requested
`
`by a user, and has invested $1.2 billion in Loudoun County, VA, including investments in the
`
`construction and maintenance of multiple data centers. (See id.;
`
`https://www.google.com/about/datacenters/locations/loudoun-county/).
`
`13.
`
`14.
`
`Google has previously submitted to the jurisdiction of this Court.
`
`Venue is proper pursuant to 28 U.S.C. §§ 1391 and/or 1400(b), at least because
`
`Google has committed acts of infringement in this judicial district and has a regular and
`
`established places of business in this judicial district.
`
`
`
`4
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 5 of 23 PageID# 5
`
`THE ASSERTED PATENTS
`
`U.S. Patent No. 9,031,259
`
`15.
`
`On May 12, 2015, the USPTO duly and legally issued United States Patent No.
`
`9,031,259 (“the ’259 patent”) entitled “Noise Reduction Apparatus, Audio Input Apparatus,
`
`Wireless Communication Apparatus, and Noise Reduction Method” to inventor Takaaki
`
`Yamabe.
`
`16.
`
`17.
`
`18.
`
`The ’259 patent is presumed valid under 35 U.S.C. § 282.
`
`SoundClear owns all rights, title, and interest in the ’259 patent.
`
`SoundClear has not granted Defendant an approval, an authorization, or a license
`
`to the rights under the ’259 patent.
`
`19.
`
`The ’259 relates to, among other things, “a noise reduction method that can
`
`reduce a noise component varied by a voice signal in a variety of environments.” ’259 patent,
`
`Col. 1, lines 41-42.
`
`20.
`
` The method determines “whether or not a sound … is a speech segment.” Id. at
`
`Col. 2, lines 18-20. When determining that the sound is the speech segment, “a voice incoming
`
`direction indicating from which direction a voice sound travels” is detected. Id. at 20-22, 25-27.
`
`A noise reduction process is then performed based on “speech segment information” and “voice
`
`incoming-direction information.” Id. at 27-32.
`
`21.
`
`This manner of reducing noise also allows for transmission of high quality voice
`
`sound even in “an environment of high noise level.” Id., Col. 1, lines 29-24; see also, e.g., Col.
`
`52, lines 60-62.
`
`
`
`5
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 6 of 23 PageID# 6
`
`U.S. Patent No. 9,070,374
`
`22.
`
`On June 30, 2015, the USPTO duly and legally issued United States Patent No.
`
`9,070,374 (“the ’374 patent”) entitled “Communication Apparatus and Condition Notification
`
`Method for Notifying a Used Condition of Communication Apparatus by Using a Light-Emitting
`
`Device Attached to Communication Apparatus” to inventors Masaya Konishi and Tatsuya
`
`Onoda.
`
`23.
`
`24.
`
`25.
`
`The ’374 patent is presumed valid under 35 U.S.C. § 282.
`
`SoundClear owns all rights, title, and interest in the ’374 patent.
`
`SoundClear has not granted Defendant an approval, an authorization, or a license
`
`to the rights under the ’374 patent.
`
`26.
`
`The ’374 relates to, among other things, “a condition notification method that
`
`achieve[s] transmission of clear voice sounds with an effective noise-cancellation function.” ’374
`
`patent, Col. 1, lines 56-58.
`
`27.
`
`The method switches a “communication mode between a standby mode … and a
`
`transmission mode.” Id. at Col. 2, lines 11-15.
`
`28.
`
`29.
`
`The method “determine[s] a pick-up state of the voice sound.” Id., lines 16-17.
`
`The method “control[s] the light-emitting device … based on the communication
`
`mode … and the pick-up state of the voice sound.” Id., lines 19-24.
`
`30.
`
`31.
`
`The method also “evaluate[s] speech quality of a speech signal.” Id., lines 31-32.
`
`This manner of reducing noise also allows for communication with clear voice
`
`sounds even in “noisy environments.” Id., Col. 1, lines 25-28.
`
`
`
`6
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 7 of 23 PageID# 7
`
`U.S. Patent No. 9,804,819
`
`32.
`
`On October 31, 2017, the USPTO duly and legally issued United States Patent
`
`No. 9,804,819 (“the ’819 patent”) entitled “Receiving Apparatus and Control Method” to
`
`inventors Kazuomi Tachigi, Kanji Kuroiwa, and Hiroshi Nakamura.
`
`33.
`
`34.
`
`35.
`
`The ’819 patent is presumed valid under 35 U.S.C. § 282.
`
`SoundClear owns all rights, title, and interest in the ’819 patent.
`
`SoundClear has not granted Defendant an approval, an authorization, or a license
`
`to the rights under the ’819 patent.
`
`36.
`
`The ’819 relates to, among other things, “a volume controller configured to
`
`cause” the output of audio “having a volume level corresponding to the operating value in a non-
`
`locked state … and … corresponding to a lock value in a locked state.” ’819 patent, Col. 2, lines
`
`9-16.
`
`37.
`
`This manner of overcoming noise in the environment also prevents the output of
`
`“an unintentional volume level … as soon as the locked state is canceled” when “the volume is
`
`reset after the volume has been locked.” Id., Col. 1, lines 30-40.
`
`BACKGROUND OF THE INVENTIONS
`
`38.
`
`These patents have been generated by the R&D engineers of a major audio
`
`processing product power house, namely JVC, now known as JVC Kenwood (“JVCK”).
`
`39.
`
`JVCK is well known for producing quality, leading-edge audio and associated
`
`products and has a long and esteemed history in doing so.
`
`40.
`
`The Patents-in-Suit were developed within the R&D department of JVCK, which
`
`consisted of many thousands of professional engineers spread over a number of R&D facilities.
`
`
`
`7
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 8 of 23 PageID# 8
`
`41.
`
`Over the years, JVCK employed the host of audio technologies that it developed
`
`to bring forward an array of leading-edge products to market.
`
`42.
`
`JVCK typically invested $260m in R&D per year to develop commercially viable
`
`technologies capable of generating substantial revenues.
`
`43.
`
`JVCK has, for various reasons, realigned its technology focus over recent years,
`
`which has led the company to divest a number of patents it developed.
`
`44.
`
`SoundClear has acquired these patents and has worked to identify companies that
`
`it believes are utilizing the technologies and profiting from the claimed inventions.
`
`CLAIMS FOR RELIEF
`
`COUNT I - Infringement of the ’259 patent
`
`45.
`
`SoundClear repeats, realleges, and incorporates by reference, as if fully set forth
`
`here, the allegations of the preceding paragraphs above.
`
`46.
`
`On information and belief, Defendant (or those acting on their behalf) make, use,
`
`sell, sell access to, import, offer to sell and/or offer to sell access to the Google Products and
`
`Services in the United States that infringe (literally and/or under the doctrine of equivalents) at
`
`least claim 1 of the ‘259 patent.
`
`47.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a noise reduction apparatus (e.g., a sound processing method in
`
`one or more of the Google Home Products that perform noise reduction by filtering noise from
`
`received audio signals.)
`
`48.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a speech segment determiner (e.g., computer
`
`hardware/software for performing signal processing (e.g., “keyword spotting” algorithm,
`
`
`
`8
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 9 of 23 PageID# 9
`
`endpointing algorithm, and/or an ASR algorithm)) configured to determine whether or not a
`
`sound (e.g., any sound in the area of one or more of the Google Home Products) picked up by at
`
`least either a first microphone or a second microphone (e.g., MEMS microphones) is a speech
`
`segment (e.g., a “keyword,” “trigger phrase,” voice request that follows the keyword/trigger
`
`phrase) and to output speech segment information (e.g., details/parameters of the speech segment
`
`itself) when it is determined that the sound picked up by the first or the second microphone is the
`
`speech segment.
`
`49.
`
`On information and belief and as an example, one or more components of the
`
`Google Products and Services include two microphones, such as the
`
`InvenSense INMP621 MEMS microphones, that form a microphone array, which receives audio
`
`from one or more sound sources including the voice of a person. See
`
`https://www.ifixit.com/Teardown/Google+Home+Teardown/72684.
`
`50.
`
`On information and belief, one or more components of the Google Products and
`
`Services determines whether sounds that are picked up are speech segments. See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/46554.pdf
`
`(Google Home Products use “endpointing” to “detect quickly and accurately when the user
`
`started and finished speaking their query.”)
`
`51.
`
`On information and belief, one or more components of the Google Products and
`
`Services outputs speech segment information (e.g., details/parameters of the speech segment
`
`itself) when it is determined that the sound picked up by the first or the second microphone is the
`
`speech segment. See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/46554.pdf; see
`
`
`
`9
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 10 of 23 PageID# 10
`
`also https://storage.googleapis.com/gweb-research2023-
`
`media/pubtools/pdf/bbd259619fdae1b2e8108348753c91c4a8f081fd.pdf.
`
`52.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a voice direction detector (e.g., using an algorithm for detecting
`
`the location of the voice audio source (e.g., “neural network adaptive beamforming (NAB)”))
`
`configured, when receiving the speech segment information, to detect a voice incoming direction
`
`(e.g., direction from voice audio source location) indicating from which direction a voice sound
`
`(e.g., audio sound of a voice) travels, based on a first sound pick-up signal (e.g., x1(k)[t])
`
`obtained based on a sound picked up by the first microphone (e.g., a sound on a first Mic Input
`
`signal) and a second sound pick-up signal (e.g., x2(k)[t]) obtained based on a sound picked up by
`
`the second microphone (e.g., a sound on a second Mic Input signal) and to output voice
`
`incoming-direction information (e.g., “frequency responses of the predicted beamforming filters
`
`at the target speech…direction[]”) when the voice incoming direction is detected. See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/45399.pdf.
`
`53.
`
`On information and belief and as an example, one or more components of the
`
`Google Products and Services determines that a sound picked up by one of its microphones is a
`
`speech segment.
`
`54.
`
`On information and belief, one or more components of the Google Products and
`
`Services also detects a voice incoming direction indicating from which direction a voice sound
`
`travels. For example, one or more of the Google Home Products include and/or utilize hardware
`
`and software that perform a “neural beamforming” technique which determines the location of
`
`the audio source. See https://www.youtube.com/watch?v=vWLcyFtni6U at 5:29-5:49. Google’s
`
`neural beamforming technique is also called “neural network adaptive beamforming (NAB).”
`
`
`
`10
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 11 of 23 PageID# 11
`
`See https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/45399.pdf.
`
`The NAB process is based on voice incoming-direction information (e.g., “frequency responses
`
`of the predicted beamforming filters at the target speech…direction[]”) and the speech segment
`
`information (e.g., details/parameters of the speech segment itself). See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/45399.pdf.
`
`55.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a noise reduction apparatus comprising an adaptive filter
`
`configured to perform a noise reduction process (e.g., a sound processing method of one or more
`
`of the Google Home Products that performs noise reduction by filtering noise from received
`
`audio signals) using the first and second sound pick-up signals based on the speech segment
`
`information (e.g., details/parameters of the speech segment itself) and the voice incoming-
`
`direction information (e.g., “frequency responses of the predicted beamforming filters at the
`
`target speech…direction[]”). See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/45399.pdf.
`
`56.
`
`On information and belief, Defendant directly infringe at least claim 1 of the ’259
`
`patent in violation of 35 U.S.C. § 271(a) by making, using, selling, selling access to, importing,
`
`offering for sale, and/or offering to sell access to the Google Products and Services.
`
`57.
`
`Defendant’ infringement has damaged SoundClear and caused / continues to
`
`cause it to suffer irreparable harm and damages.
`
`COUNT II - Infringement of the ’374 patent
`
`58.
`
`SoundClear repeats, realleges, and incorporates by reference, as if fully set forth
`
`here, the allegations of the preceding paragraphs above.
`
`
`
`11
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 12 of 23 PageID# 12
`
`59.
`
`On information and belief, Defendant (or those acting on their behalf) make, use,
`
`sell, sell access to, import, offer to sell and/or offer to sell access to the Google Products and
`
`Services in the United States that infringes (literally and/or under the doctrine of equivalents) at
`
`least claim 9 of the ’374 patent.
`
`60.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a communication apparatus, such as the Google Home
`
`Products.
`
`61.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a first pick-up unit (e.g., a first microphone) configured to pick
`
`up a voice sound (e.g., a spoken sound originating from a person (e.g., a voice request and/or
`
`wake word)). See e.g., https://www.ifixit.com/Teardown/Google+Home+Teardown/72684; See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/46554.pdf.
`
`62.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a transmitter unit (e.g., unit in one or more of the Google
`
`Home Products that sends the trigger phrase and/or voice request to Google’s secure server)
`
`configured to transmit the voice sound picked up by the first pick-up unit to outside (e.g.,
`
`Google’s secure server) as a first speech signal (e.g., data signal representing the trigger phrase
`
`and/or voice request).
`
`63.
`
`On information and belief, one or more components of the Google Products and
`
`Services uses a client-server keyword spotting system (“KWS”) triggered by a keyword or key
`
`phrase such as “Ok, Google.” See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/46554.pdf.
`
`
`
`12
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 13 of 23 PageID# 13
`
`64.
`
`On information and belief, Google Home Products transmit the voice sound (e.g.,
`
`the keyword and/or the phrases after the keyword) to the server via a transmitter. See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/46554.pdf.
`
`65.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides comprise a communication-mode switching unit (e.g., unit in
`
`one or more of the Google Home Products that causes the transmitter unit to send or not send the
`
`trigger phrase and/or voice request to Google’s secure server) configured to switch a
`
`communication mode (e.g., operation mode of the transmitter unit) between a standby mode
`
`(e.g., when the microphones are on but the trigger phrase is not detected) in which the transmitter
`
`unit does not transmit the speech signal and a transmission mode (e.g., when microphones are on
`
`and the trigger phrase is detected) in which the transmitter unit transmits the speech signal.
`
`66.
`
`On information and belief, the processor in the one or more components of the
`
`Google Products and Services transmits the first speech signal when the keyword is detected
`
`using the “keyword spotting” feature. See
`
`https://www.ifixit.com/Teardown/Google+Home+Teardown/72684; See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/46554.pdf.
`
`67.
`
`On information and belief, one or more components of the Google Products and
`
`Services controls the transmitter unit to transmit the first speech signal when the keyword/trigger
`
`phrase (e.g., “OK Google”) is detected using the keyword spotting feature. See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/46554.pdf
`
`(“Once a trigger phrase is detected on device, typically the connection is opened to the server
`
`and the audio corresponding to the rest of the query (e.g. ‘play some music’) is sent for
`
`
`
`13
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 14 of 23 PageID# 14
`
`transcription using a server-side [automatic speech recognition] ASR system.” See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/46554.pdf. .
`
`68.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a sound pick-up state determination unit (e.g., (1)
`
`hardware/software for performing audio signal processing and filtering (e.g., “neural network
`
`adaptive beamforming (NAB)”), and (2) computer hardware/software for performing a speech
`
`recognition algorithm (e.g., a “keyword spotting” algorithm and/or an ASR algorithm))
`
`configured to determine a pick-up state (e.g., signal characteristics) of the voice sound picked up
`
`by the first pick-up unit. See https://www.youtube.com/watch?v=vWLcyFtni6U;
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/45399.pdf;
`
`https://support.google.com/assistant/answer/9712065?sjid=16295118774734419352-AP.
`
`69.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a light emission device (e.g., an LEDs) configured to emit
`
`light. See https://support.google.com/googlenest/answer/7073219?hl=en#zippy=%2Cgoogle-
`
`home.
`
`70.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a control unit (e.g., unit in one or more of the Google Home
`
`Products that controls the LED lights to operate in a variety of modes) configured to control the
`
`light-emitting device so that the light-emitting device is turned off, turned on or repeatedly
`
`turned on and off (e.g., “spin,” “continuously spin,” “pulse,” “continuously pulse,” “solid”)
`
`based on the communication mode switched by the communication-mode switching unit, and the
`
`pick-up state of the voice sound picked up by the first pick-up unit and determined by the sound
`
`pick-up state determination unit.
`
`
`
`14
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 15 of 23 PageID# 15
`
`71.
`
`On information and belief, one or more components of the Google Products and
`
`Services controls the LED ring light to operate in a variety of modes based on the pick-up state
`
`of the voice sound. For example, the devices include two “NXP PCA9956BTW LED drivers”
`
`and one “Atmel ATSAMD21 32-bit ARM Cortext-M0+ microcontroller.” See
`
`https://www.ifixit.com/Teardown/Google+Home+Teardown/72684. Together these components
`
`form a unit that controls the operational states of the LEDs based on other functions and
`
`operations of the device (e.g., the communication mode and the pick-up state of the voice
`
`sound). Id.
`
`72.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a speech-quality evaluation unit (e.g., hardware/software for
`
`performing audio signal processing and filtering (e.g., “neural network adaptive beamforming
`
`(NAB),” “‘Hey Google’ sensitivity”)) configured to evaluate speech quality (e.g., determine the
`
`degree of signal processing/filtering necessary to produce an “enhanced speech signal,” and
`
`produce an enhanced speech signal) of the first speech signal to be transmitted by the transmitter
`
`unit, wherein the sound pick-up state determination unit determines the sound pick-up state of
`
`the voice sound picked up by the first sound pick-up unit based on the speech quality of the
`
`speech signal evaluated by the speech-quality evaluation unit.
`
`73.
`
`On information and belief, one or more of the Google Products and Services
`
`includes hardware/software for performing a “neural beamforming” technique which increases
`
`the quality and accuracy of speech recognition while only using two microphones. See
`
`https://www.youtube.com/watch?v=vWLcyFtni6U at 5:29-5:49. Google’s neural beamforming
`
`technique is also called “neural network adaptive beamforming (NAB).” See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/45399.pdf.
`
`
`
`15
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 16 of 23 PageID# 16
`
`74.
`
`On information and belief, Google’s NAB process uses “[Long Short-Term
`
`Memory] LSTM layers to predict time domain beamforming filter coefficients at each input
`
`frame,” thereby allowing the algorithm to “adapt to previously unseen or changing conditions.”
`
`Id. at 1. The NAB process evaluates speech quality by adjusting filter coefficients on a frame-by-
`
`frame basis of the received audio signal. Id. (The NAB model “re-estimates a set of spatial filter
`
`coefficients at each input frame using a neural network. Specifically, raw multichannel
`
`waveform signals are passed into a filter prediction (FP) LSTM whose outputs are used as spatial
`
`filter coefficients. These spatial filters for each channel are then convolved with the
`
`corresponding waveform input, and the outputs are summed together to form a single channel
`
`output waveform containing the enhanced speech signal.”). Id.
`
`75.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a speech-segment determination unit (e.g., computer
`
`hardware/software for performing a speech recognition algorithm (e.g., a “keyword spotting”
`
`algorithm, endpointing algorithm, and/or a ASR algorithm)) configured to determine whether or
`
`not the first speech signal to be transmitted by the transmitter unit is a speech segment (e.g., a
`
`“keyword,” “trigger phrase,” voice request that follows the keyword/trigger phrase), wherein,
`
`the sound pick-up state determination unit determines the sound pick-up state of the sound to be
`
`transmitted as the first speech signal based on a determination result (e.g., a “spotted” or
`
`“detected” keyword/trigger phrase) at the speech-segment determination unit and an evaluation
`
`result (e.g., “enhanced speech signal”) at the speech-quality evaluation unit.
`
`76.
`
`On information and belief, one or more components of the Google Products and
`
`Services includes a client-server KWS (keyword spotting system) triggered by a keyword, e.g.,
`
`“OK google.” See
`
`
`
`16
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 17 of 23 PageID# 17
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/46554.pdf.
`
`The products include a transmitter unit (e.g., a Marvell Avastar 88W8887 WLAN/BT/NFC SOC
`
`(see chip in yellow box below)) which communicates with a server in the KWS system. See
`
`https://www.ifixit.com/Teardown/Google+Home+Teardown/72684; see also
`
`https://support.google.com/googlenest/answer/7072284?hl=en&sjid=18139057093911631537-
`
`NA#zippy=%2Cgoogle-home. Google Home Products transmit the voice sound (e.g., the
`
`keyword and/or the phrases after the keyword) to the server via a transmitter. See also See
`
`https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/46554.pdf.
`
`77.
`
`On information and belief, Defendant directly infringe at least claim 9 of the ‘374
`
`patent in violation of 35 U.S.C. § 271(a) by making, using, selling, selling access to, importing,
`
`offering for sale, and/or offering to sell access to the Google Products and Services.
`
`78.
`
`Defendant’ infringement has damaged SoundClear and caused / continues to
`
`cause it to suffer irreparable harm and damages.
`
`COUNT III - Infringement of the ’819 patent
`
`79.
`
`SoundClear repeats, realleges, and incorporates by reference, as if fully set forth
`
`here, the allegations of the preceding paragraphs above.
`
`80.
`
`On information and belief, Defendant (or those acting on their behalf) make, use,
`
`sell, sell access to, import, offer to sell and/or offer to sell access to the Google Products and
`
`Services in the United States that infringes (literally and/or under the doctrine of equivalents) at
`
`least claim 8 of the ’819 patent.
`
`81.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a control method (e.g., a method for regulating and processing
`
`
`
`17
`
`

`

`Case 1:24-cv-00729 Document 1 Filed 05/01/24 Page 18 of 23 PageID# 18
`
`sound signals) for a receiving apparatus (e.g., one or more of the Google Home Products). See
`
`https://www.ifixit.com/Teardown/Google+Home+Teardown/72684.
`
`82.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a control method for a receiving apparatus comprising an audio
`
`output unit (e.g., a speaker) configured to output audio (e.g., the sound from a speaker of Google
`
`Assistant’s spoken responses) corresponding to an audio signal (e.g., input signal to the speaker
`
`representing Google Assistant’s spoken responses). See
`
`https://www.ifixit.com/Teardown/Google+Home+Teardown/72684.
`
`83.
`
`On information and belief, one or more components of the Google Products and
`
`Services is, employs, or provides a control method for a receiving apparatus comprising a
`
`volume operating unit (e.g., software in one or more of the Google Home Products that controls
`
`the volume of Google Assistant’s spoken responses) configured to output an operating value
`
`(e.g., an electronic signal va