Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 1 of 67
`
`UNITED STATES DISTRICT COURT
`WESTERN DISTRICT OF TEXAS
`
`WACO DIVISION
`CELLULAR SOUTH, INC. §
`§
`Plaintiff, §
`§ :
`v. § Civil Action No. 6:24-¢v-00245
`§
`GOOGLE, LLC §
`§ JURY TRIAL DEMANDED
`Defendant. §
`§
`§
`§
`§
`
`COMPLAINT FOR PATENT INFRINGEMENT
`DEMAND FOR JURY TRIAL
`
`Plaintiff Cellular South Inc. (“CSI” or “Plaintiff”) files this Complaint against defendant
`Google, LLC (“Google” or “Defendant”) and alleges as follows:
`NATURE OF THIS ACTION
`
`1. This complaint alleges patent infringement. CSI alleges that Google has infringed
`and continues to infringe three patents: U.S. Patent Nos. 10,218,954 (“the *954 Patent”), 9,940,972
`(“the 972 Patent”), and 11,126,853 (“the *853 Patent”). Copies of these patents (collectively, the
`“Patents-in-Suit”) are attached hereto as Exhibits A—C.
`
`2. The Patents-in-Suit cover foundational technologies for organizing unstructured
`data within video that allows content providers to gain a better understanding of the context and
`value of content and present it back to users in a highly personalized manner.
`
`3. Google directly infringes the Patents-in-Suit by making, using, offering to sell,
`selling, and/or importing into the United States video classification and recognition technology,
`software, and services that practice the inventions claimed in the Patents-in-Suit. Google directs
`and controls each relevant aspect of the accused technology discussed herein, and benefits from
`
`the use of each feature that infringes the Patents-in-Suit.
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 1 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 2 of 67
`
`4. CSI seeks damages and other relief for Google’s infringement of the
`Patents-in-Suit.
`
`THE PARTIES
`
`5. Plaintiff CSI is a corporation formed under the laws of Mississippi with a principal
`place of business at 1018 Highland Colony Parkway, Suite 300, Ridgeland, Mississippi 39157.
`
`6. Defendant Google LLC is a corporation formed under the laws of Delaware with a
`principal place of business at 1600 Amphitheatre Parkway, Mountain View, California 94043.
`Google also has an office located at 500 West 2nd Street in Austin, Texas.
`
`JURISDICTION AND VENUE
`
`7. CSI brings this civil action for patent infringement under the Patent Laws of the
`United States, 35 U.S.C. § 1 et. seq., including 35 U.S.C. §§ 271, 281-285. This Court has subject
`matter jurisdiction over this action under 28 U.S.C. §§ 1331 and 1338.
`
`8. Venue is proper in this judicial district pursuant to 28 U.S.C. § 1400(b). Google
`maintains an established place of business in the state of Texas, and the Western District of Texas
`
`specifically, including its office at 500 West 2nd Street, Austin, Texas 78701.
`
`= GoogleCareers Teoms Locatons Benefits Jobs _ Students Your career =
`
`About Go gle v About GoogleintheUS. Products Commitments Stories The Keyword
`
`North America Latin America
`
`Ann Arbor
`
`% 2300 Traverwood Dr.
`Ann Arbor, MI 48105
`United States
`Phone: +1734-332-6500
`Directions
`
`Austin
`
`115 jobs available
`
`Austin
`
`500 W 2nd St
`
`Suite 2900
`
`Austin, TX 78701
`
`United States
`
`Phone: +1 512-343-5283
`Directions
`
`Source:https://www.google.com/about/careers/applications/locations/austin/ (last accessed Mar.
`1, 2024); https://about.google/locations/?region=north-america (last accessed Mar. 1, 2024).
`
`Additionally, on information and belief, Google has committed acts of infringement in this judicial
`
`district, and has purposefully transacted business in this judicial district.
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 2 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 3 of 67
`
`9. Google sells and offers for sale the accused technology in this state, including
`within the Western District of Texas. More specifically, Google offers the ability for anyone in the
`state to purchase its Cloud Video Intelligence and Video Al software platform that includes the
`accused technology. Additionally, upon information and belief, Google employs personnel who
`work on and use the accused technology in their Austin, Texas office. Google, therefore, obtains
`the benefits and protections of the laws of the State of Texas. This dispute arises out of and has a
`substantial connection with Google’s contacts within this state and its infringement in this state,
`resulting in the exercise of jurisdiction being fair and reasonable.
`
`10. Google is subject to this Court’s specific and general personal jurisdiction pursuant
`to due process and/or the Texas Long Arm Statute because Google conducts substantial business
`in this forum, including: (i) making, using, selling, importing, and/or offering for sale the accused
`technology all throughout the District; and (ii) upon information and belief, employs engineers
`and other personnel who work on, sell, and use the accused technology.
`
`FACTUAL BACKGROUND
`
`11. CSI is a diversified telecommunications and technology company based in
`Ridgeland, Mississippi. CSI’s mission is to “engage the exceptional and embrace operational
`excellence to best deliver connectivity and technology solutions that advance our communities and
`customers’ lives.”! CSI employs approximately 1,850 people who work and live in Mississippi,
`Alabama, and Tennessee. It has three lines of business—Wireless, Home Fiber and Business.
`
`12. CSl s part of a family of operating companies owned by a privately held company,
`Telapex, Inc. Telapex, Inc. is owned by the Creekmore family, which has been in the
`telecommunications business in Mississippi since 1948.
`
`13. Starting in the 1948, Wade Creekmore, Sr. began to acquire small rural telephone
`exchanges in areas of Mississippi which Southern Bell had no interest in serving. In the late
`
`1950’s, Mr. Creekmore divided the properties into two operating landline telephone companies—
`
`! Cellular South Corporate Mission Statement. See https://www.cspire.com/web/company/about.
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 3 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 4 of 67
`
`Franklin Telephone Company, Inc. and Delta Telephone Company, Inc.—and put his sons Wade,
`Jr. and Jimmy in charge of each, respectively. As incumbent telephone operators, Franklin and
`Delta were eligible to participate in the first lotteries conducted by the Federal Communications
`Commission, and in 1986, they acquired their first wireless licenses in the State of Mississippi.
`Two years later, CSI launched the first 1G wireless service on the Mississippi Gulf Coast. Within
`the next decade, CSI brought wireless telephone services to forty-two (42) Mississippi counties.
`
`14. Beginning in the early 2000s, CSI expanded its footprint to include parts of
`Tennessee and northern Mississippi, upgraded to 4G service, and then to 5G. CSI remains a true
`regional wireless provider and remains committed to its community through the continued
`expansion of telecommunications technology and its investment into the education and
`development of the Mississippi community.
`
`15. In addition to its wireless business, CSI through its subsidiaries owns or manages
`more than 20,000 miles of optical fiber networks primarily in the states of Mississippi and
`Alabama. CSI is also a leading value-added reseller for commercial hardware and software
`applications offered by major technology companies and solutions providers. Using its fiber and
`commercial technology solutions, CSI connects businesses and government entities with a suite of
`world-class IT solutions.
`
`16. CSI announced its Fiber to the Home initiative in 2013, when there were only four
`cities in the United States where fiber to the home was available. In 2014, CSI launched its first
`market in the small Mississippi town of Quitman, which became one of the first communities in
`the United States to have buried fiber providing speeds of 1 Gigabit per second (Gig Internet
`Service) to the home. Today CSI has residential fiber customers in about 115 communities in
`Mississippi and Alabama, and it has recently developed new “fiberhoods” in Tennessee as well.
`
`17. While CSI is a relatively small company compared to its competitors in the
`telecommunications industry, it has always been an innovator. Beginning with its founder’s
`willingness to offer telephone services in rural areas where the Bell System would not, CSI
`
`continues to lead through its commitment to offer the best wireless service using the latest
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 4 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 5 of 67
`
`technologies, as exemplified by its position as one of the first providers to offer Gig Internet
`Service over fiber to residential areas.
`
`18. CSI has also worked to diversify its product offerings to remain innovative and
`competitive in the telecommunications and technology marketplace. Through one of these efforts,
`CSI created and offered “Video-to-Data” (or “V2D”), a cloud-based digital profiling and analytics
`system that could scan and turn video into searchable data faster than the real-time viewing of the
`video by using multiple servers that run simultaneously and in parallel. CSI offered V2D
`commercially through a wholly-owned subsidiary named Vi Digital, LLC (“Vu Digital” or “Vu”).
`
`19. V2D breaks down a video frame-by-frame and identifies the objects—for example,
`text, audio, images, faces, locations, logos—within each frame. V2D then creates a chronological
`record of all the objects identified within the video. V2D was a transformative product that made
`video content as easy to search as text, thereby providing users with unprecedented video
`classification and clustering capabilities, as well as significantly enhanced search engine indexing,
`content personalization, and targeted advertising capabilities. As Vu Digital’s spokesperson
`explained: “We [Vu] offer the only single, comprehensive automated solution on the market today
`capable of organized the unstructured data within video,” that will help content providers “unlock
`the power of video.”?
`
`20. CSI believed V2D could be an incredibly useful technology for video data
`processing and content personalization and set out to market it to a broad variety of potential
`customers. By at least 2016, Vu Digital outlined how V2D could be used by law enforcement to
`process body camera footage which was becoming more widely adopted around this same time.
`
`Vu Digital also marketed V2D to a variety of entertainment, technology, and sports vendors, many
`
`2 https://www.cspire.com/cms/news/wireless/25500006/V%C5%AB%20Digital%20provides
`%20PGA%20TOUR%20Entertainment%20Division%20with%20Video-to-
`Data%20(V2D)%20Analytics%20Services (last accessed Mar. 5, 2024).
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 5 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 6 of 67
`
`of whom expressed interest in the product. For example, in or around 2016, Vu entered a multi-
`year agreement with the PGA Tour to help process the PGA’s video and audio content.’
`
`21. The V2D product debuted in May 2015 and mere months later was recognized by
`TCM—a global integrated media conglomerate and leading source of news and information for
`the communications and technology industries worldwide—as the “best-of-the-best technology
`solutions available on the market today.”* V2D received TCM’s coveted Communications
`Solutions 2015 Product of the Year Award that recognizes exceptional voice, data, and video
`communications products and services. Other 2015 recipients for this honor included prominent
`industry players such as AT&T, Alcatel-Lucent, Comcast, Dell, Hewlett Packard, and Logitech.’
`
`22. Vu’s V2D product also drew praise from Innovate Mississippi, a statewide group
`that helps innovation-based startup companies by connecting entrepreneurs with investors. As
`Innovate’s then-President and CEO Tony Jeff explained, “It’s not every day that a Mississippi
`company is mentioned in the same breath alongside some of the world’s technology giants, but
`thanks to [CSI] and Vii Digital, it’s becoming more common.”®
`
`23. In addition to the TMC award, Vu Digital was selected in 2015 to be part of the
`National Association of Broadcasters’ (“NAB”) year-long SPROCKIT accelerator program for
`innovative startups in media and entertainment.” SPROCKIT is a global innovation platform
`created to help large media, entertainment, and technology companies meet with emerging tech
`
`start-ups to fast-track investment, acquisition, and partnerships between the two groups on new
`
`31d.
`
`4 https://www.cspire.com/cms/news/wireless/24600006/V%C5%AB%20Digital%20Wins%20
`Coveted%20TMC%20Communications%20Solutions%202015%20Product%200f%20the%2
`0Year%20Award (last accessed Mar. 5, 2024).
`
`> Id.
`°Id.
`7Id.
`IPR2025-00877
`Patent Owner Exhibit 2008
`
`Page 6 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 7 of 67
`
`products and services. On information and belief, Vu Digital’s Wade Smith presented at the
`SPROCKIT NAB Show, which was held April 11-16, 2015 in Las Vegas.®
`
`24. SPROCKIT Sync is invitation-only event put on three times a year by SPROCKIT
`that entails a series of private meetings between selected startups and corporate players in the
`media and entertainment industry. Following the launch of its V2D product, Vu Digital was
`invited to attend the SPROCKIT Sync conference that was held on June 18, 2015 at Google Tech
`Corners in Sunnyvale, California. Vu Digital’s Project Specialist, Gregory Sandifier, presented
`its then “patent pending” V2D technology at the conference. On information and belief, when Vu
`Digital presented at this conference in 2015, the corporate partners participating included Disney,
`
`Fox, Samsung, and Google, among others. Vu Digital presented V2D to these corporate partners,
`
`including Google.
`FIRST CLAIM FOR RELIEF
`(Infringement of U.S. Patent. No. 10,218,954)
`25. CSI realleges and incorporates by reference the allegations of the foregoing
`paragraphs.
`
`26. On February 26, 2019, the United States Patent and Trademark Office (“USPTO”)
`duly and legally issued, after a full and fair examination, United States Patent No. 10,218,954 (the
`“’954 Patent”) entitled “Video to Data” to inventors Nacem Lakhani, Bartlett Wade Smith, IV,
`and Allison A. Talley. A true and correct copy of the *954 patent is attached as Exhibit A to this
`Complaint.
`
`27. The °954 Patent was assigned to CSI, which currently holds all substantial rights,
`title, and interest in and to the 954 Patent.
`
`28. Pursuant to 35 U.S.C. § 282, the *954 Patent is presumed valid.
`
`29. The *954 Patent is presumed to be patent eligible under 35 U.S.C. § 101.
`
`8 https://www.businesswire.com/news/home/20150402006447/en/NAB-Show-Unveils-Final-10-
`Participants-Selected-for-SPROCKIT-2015 (last accessed Mar. 5, 2024).
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 7 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 8 of 67
`
`30. The ’954 Patent is directed to an improvement in the functionality of machine
`learned video recognition and classification systems, particularly with regard to specific
`techniques for improving the accuracy of predictions made using image, audio, text, and other
`video data. More specifically, the 954 Patent is directed to techniques that utilize contextual
`information such as the arrangement of certain objects in a series of still images in order to improve
`computer visual recognition in a video classification system.
`
`31. The ’954 Patent addresses specific technological challenges that arose in video
`classification systems when attempting to extract meaningful metadata from video content.
`
`32. The 954 Patent describes the challenges in the field of computer visual recognition
`
`systems and explains the advantages of the claimed inventions:
`
`Identifying various objects in an image can be a difficult task. For
`example, locating (segmenting) and positively identifying an object
`in a given frame or image can yield false positives-locating but
`wrongfully identifying an object. Therefore, present embodiments
`can be utilized to eliminate false positives, for example, by using
`context. As one example, if the audio soundtrack of a video is an
`announcer calling a football game, then identification of ball in a
`given frame as basketball can be assigned a reduced probability or
`weighting. As another example of using context, if a given series of
`image frames from a video is positively or strongly identified as a
`horse race, then identifying an object to be a mule or donkey can be
`given a reduced weight.’
`
`33. The ’954 Patent provides additional examples illustrating the improvements that
`
`the described methodologies provide to computer visual recognition systems:
`
`In certain instances, identification of an individual can be a difficult
`task. For example, facial recognition can become difficult when an
`individual’s face is obstructed by another object like a football, a
`baseball helmet, a musical instrument, or other obstructions. An
`advantage of some embodiments described herein can include the
`ability to identify an individual without identification of the
`individual’s face. Embodiments can use contextual information
`such as association of objects, text, and/or other context within an
`
`97954 Patent, 5:3—16.
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 8 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 9 of 67
`
`image or video. As one example, a football player scores a
`touchdown but rather than identifying the player using facial
`recognition, the play can be identified by object recognition of, for
`example, the player’s team’s logo, text recognition of the player’s
`jersey number, and by cross referencing this data with that team’s
`roster (as oppose to another team, which is an example of why the
`logo recognition can be important). Such embodiments can further
`learn to identify that player more readily and save his image as
`data.'®
`
`34. The ’954 Patent explains that the described methodologies for deriving contextual
`information can be helpful in identifying and correcting or eliminating false positives. !!
`Moreover, these methodologies may make it possible to identify unknown objects in a given image
`by narrowing a large number of practically infinite possibilities to a relatively small number of
`object possibilities, thereby allowing positive object recognition where identification previously
`could not be achieved.'?
`
`35. As the examiner acknowledged during prosecution, none of the prior art cited
`during examination disclosed the claimed techniques which recite using audio and video semantic
`information to generate “contextual topics” and “generating a contextual text, an image, or an
`animation” based on the determined contextual topics.
`
`36. The technological solutions described above are recited in the *954 Patent claims,
`
`including, for example, independent claims 1 and 13 (and their corresponding dependent claims).
`
`37. Claim 1 of the 954 Patent reads as follows:
`
`1. A method to generate video data from a video comprising:
`generating audio files and image files from the video;
`
`distributing the audio files and the image files across a plurality of
`processors and processing the audio files and the image files in
`parallel;
`
`10 74 at 5:35-52.
`14 at 5:53-63.
`12 1d. at 5:64-6:14.
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 9 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 10 of 67
`
`converting audio files associated with the video to text;
`identifying an object in the image files;
`determining a contextual topic from the image files;
`
`assigning a probability of accuracy to the identified object based on
`the contextual topic;
`
`converting the image files associated with the video to video data,
`wherein the video data comprises the object, the probability, and the
`contextual topic;
`
`cross-referencing the text and the video data with the video to
`determine contextual topics;
`
`generating a contextual text, an image, or an animation based on the
`determined contextual topics;
`
`generating a content-rich video based on the generated text, image,
`or animation.
`
`38. Upon information and belief, Google directly infringes at least claim 1, at least in
`the exemplary manner described below.
`
`39. Google directly infringes the 954 Patent by making, using, offering to sell, and/or
`selling in the United States its Cloud Video Intelligence platform (“Video Intelligence” or the
`“Accused Product”), which relies on machine learning (“ML”) models and natural language
`processing (“NLP”) techniques to identify and classify videos using contextual information.
`Google directs and controls each relevant aspect of the accused technology discussed herein, and
`benefits from the use of each feature that infringes the 954 Patent. On information and belief
`Google uses the 954 patented inventions and the Accused Product to classify videos uploaded to
`
`its online video sharing platform, YouTube (www.youtube.com) and to its Google Photos i1OS and
`
`Android apps.
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 10 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 11 of 67
`
`You can now try AutoML for video data on our new complete, cohesive ML platform, Vertex Al. Learn more.
`
`Video Al Video Al
`Benefits _ . o o . CBS Interactivi
`Enable powerfu very and ging video experiences. AIML Groups
`\ storr re on V stomers =Y jo -
`ases charge SRS
`ratior
`.
`
`See how CBS Interactive
`is serving 38 media
`brands with content
`discovery.
`
`Compare features
`Pricing L Video Intelligence.
`
`Take the next step
`
`recommendations, and more
`
`Source: https://cloud.google.com/video-intelligence
`
`40. According to Google, Video Intelligence may be used with custom (via Vertex Al
`for AutoML)!® or pretrained (“Video Intelligence API”) machine learning models in order to
`classify and annotate videos so that video data may be more easily parsed or indexed and to enable
`
`users to implement, for example, contextual-based advertising and/or content recommendation for
`
`their video content libraries.'*
`
`13 On information and belief, as of January 23, 2024, Google no longer supports use of the
`“AutoML Video Intelligence” product to allow customers to classify video content using custom
`ML models. Instead, AutoML has been replaced and/or consolidated with Google’s “Vertex AI”
`product which continues to provide this capability to customers. See, e.g.,
`https://cloud.google.com/video-intelligence/automl/pricing (last accessed March 4, 2024). As
`Google explains “[a]ll of the functionality of legacy AutoML Video Intelligence and new features
`are available on the Vertex Al platform.” /d.
`
`14 See https://cloud.google.com/video-intelligence (last accessed Mar. 5, 2024).
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 11 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 12 of 67
`
`KEY FEATURES
`
`Two ways to make your media more
`discoverable and valuable
`
`AutoML Video Intelligence
`
`Vertex Al for AutoML video has a graphical interface that makes it easy to train your
`own custom models to classify and track objects within videos, even if you have
`minimal machine learning experience. It's ideal for projects that require custom
`labels that aren’t covered by the pre-trained Video Intelligence API.
`
`Video Intelligence API
`
`Video Intelligence API| has pre-trained machine learning models that automatically
`
`recognize a vast number of objects, places, and actions in stored and streaming
`video. Offering exceptional quality out of the box, its highly efficient for common
`use cases and improves over time as new concepts are introduced.
`
`Source:
`
`https://cloud.google.com/video-intelligence (last accessed Mar. 5, 2024)
`
`41. As Google explains, Video Intelligence may be used to add custom or predefined
`
`labels for both stored video content library as well as for live-streaming video applications. !°
`
`151d.
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 12 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 13 of 67
`
`Which video product is right for you?
`You can use Video Intelligence API to quickly categorize content using thousands of predefined labels or use AutoML
`Video Intelligence to create custom labels for specific needs
`AutoML Video Intelligence Video Intelligence API
`Use REST and RPC APIs
`[ (]
`Use a graphical Ul
`[
`Annotate video using predefined
`abels °
`[
`Stored video analysis
`[ (]
`[ (]
`Shot change detection
`[ (]
`Object detection and tracking
`[ [
`Text detection and extraction using
`OCR ®
`[
`Automated closed captioning and
`subtitles °
`Logo recognition
`[
`(4
`Face detection (beta)
`[
`[
`
`Source: https://cloud.google.com/video-intelligence
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 13 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 14 of 67
`
`Use cases
`
`USE CASE
`Content moderation
`
`Identify when inappropriate content is being shown in a given video. You can instantly conduct content moderation across
`petabytes of data and more quickly and efficiently filter your content or user-generated content.
`
`USE CASE
`
`Recommended content
`
`Build a content recommendation engine with labels generated by Video Intelligence APl and a user’s viewing history and
`preferences. This will simplify content discovery for your users and guide them to the most relevant content that they want.
`
`USE CASE
`Media archives
`
`Create an indexed archive of your entire video library by using the metadata from Video Intelligence API. Ideal for mass media
`companies, Video Intelligence API can automatically analyze content and make the results immediately accessible via the API.
`
`USE CASE
`Contextual advertisements
`
`You can identify appropriate locations in videos to insert ads that are contextually relevant to the video content. This can be done
`by matching the timeframe-specific labels of your video content with the content of your advertisements.
`
`Source: https://cloud.google.com/video-intelligence
`
`42. On information and belief, Google’s customers use Video Intelligence to generate
`video metadata using the context-based techniques of the 954 Patent in order to “enhance the
`
`video experience.”
`
`@ CBS Interactive
`
`“Video Intelligence allows CBS Interactive
`to plug into our existing video encoding
`framework to generate video metadata.
`The performance and reliability allows us
`
`to enhance the video experience.”
`
`Adam Leary, VP, Data Science Services, CBS Interactive
`
`Source : https://cloud.google.com/video-intelligence
`
`43. Google provides the Video Intelligence product to users through its Google Cloud
`Platform (“GCP”) and charges users for the amount of resources (e.g., cloud server virtual CPUs)
`
`consumed for video processing on a per minute basis.
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 14 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 15 of 67
`
`Feature
`
`Label detection
`
`Shot detection
`
`Explicit content detection
`
`Speech transcription
`
`Object tracking
`
`Text detection
`
`Logo detection
`
`Face detection
`
`Person detection
`
`Celebrity recognition
`
`First 1000 minutes
`
`Free
`
`Free
`
`Free
`
`Free
`
`Free
`
`Free
`
`Free
`
`Free
`
`Free
`
`Free
`
`Stored video annotation
`
`Video Intelligence API pricing
`
`Prices are per minute. Partial minutes are rounded up to the next full minute. Volume is per month.
`
`Minutes 1000+
`
`$0.10 / minute
`
`$0.05 / minute, or free with Label detection
`
`$0.10 / minute
`
`$0.048 / minute (charges for en-US transcription only)
`
`$0.15/ minute
`
`$0.15/ minute
`
`$0.15/ minute
`
`$0.10 / minute
`
`$0.10 / minute
`
`$0.10 / minute
`
`Source: https://cloud.google.com/video-intelligence/pricing
`
`Feature
`
`Label detection
`
`Shot detection
`
`Explicit content detection
`
`Object tracking
`
`Streaming video annotation
`
`First 1000 minutes Minutes 1000+
`Free $0.12 / minute
`Free $0.07 / minute
`Free $0.12 / minute
`
`Free
`
`$0.17 / minute
`
`Source: https://cloud.google.com/video-intelligence/pricing#streaming_video_annotation
`
`44. Withregard to claim 1 of the *954 Patent, Google practices “[a] method to generate
`
`video data from a video comprising: generating audio files and image files from the video.” For
`
`IPR2025-00877
`
`Patent Owner Exhibit 2008
`
`Page 15 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 16 of 67
`
`example, at the Google Next 2017 developer conference, Google explained that Video Intelligence
`
`begins by decoding video into video data, audio data, subtitle data, and various other metadata. '®
`
`Introduction to Video Intelligence (Google Cloud Next “17)
`
`' Google Cloud Tech @
`1.1M subscribers
`
`/i
`
`| Google Cloud
`
`Juhyun Lee
`
`Video Content Analyst
`Google
`
`} >l 19 20:38 / 44:04 - Separate signal from noise-frame level > ° EB ° E D @ [:
`
`Source: https://www.youtube.com/watch?v=y-k8oelbmGc¢
`
`So given a video in your GCS [Google Cloud Storage] bucket, the first thing that
`
`we do is we decode the video. And usually a video container contains like audio
`
`streams, video streams, subtitles and all of the metadata.'’
`
`45. The Accused Product performs the step of “distributing the audio files and the
`image files across a plurality of processors and processing the audio and image files in parallel.”
`
`On information and belief, the Video Intelligence product executes on a distributed network of one
`
`or more Google Cloud servers. As Google explains, Video Intelligence uses the Google Cloud
`
`16 Introduction to Video Intelligence (Google Cloud Next *17) (available at
`https://www.youtube.com/watch?v=y-k8oelbmGc) (last accessed Mar. 5, 2024).
`
`171d. at 21:00 to 22:14 (emphasis added).
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 16 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 17 of 67
`
`Platform (“GCP”), which comprises a number of computer resources housed in Google’s data
`
`centers around the world.'®
`
`Google Cloud resources
`
`together.
`
`Google Cloud consists of a set of physical assets, such as computers and hard disk drives, and virtual resources, such
`
`as virtual machines (VMs), that are contained in data centers around the globe. Each data center location is in a region.
`Regions are available in Asia, Australia, Europe, North America, and South America. Each region is a collection of zones,
`which are isolated from each other within the region. Each zone is identified by a name that combines a letter identifier
`with the name of the region. For example, zone a in the East Asia region is named asia-easti1-a.
`
`This distribution of resources provides several benefits, including redundancy in case of failure and reduced latency by
`locating resources closer to clients. This distribution also introduces some rules about how resources can be used
`
`Source: https://cloud.google.com/docs/overview
`
`(Global Scope)
`
`Static External IP Addresses
`
`Region: Central US
`
`The following diagram shows the relationship between global scope, regions and zones, and some of their resources:
`
`Google Cloud Platform
`
`Zone Zone
`
`us-central 1-a us-central 1-b
`
`{e} VMs Zone
`
`us-central 1-¢
`
`omm= Disks Zone
`us-central 1-f
`
`Source: https://cloud.google.com/docs/overview
`
`46. In addition, Google explains that a GCP account and project are required in order
`
`to use the Video Intelligence API.
`
`18 See https://cloud.google.com/docs/overview.
`
`IPR2025-00877
`Patent Owner Exhibit 2008
`Page 17 of 67
`
`
`
`
`
`
`
`
`Case 6:24-cv-00245 Document 1 Filed 05/09/24 Page 18 of 67
`
`Before you begin
`
`1. If you're new to Google Cloud, create an account to evaluate how our products perform in real-world scenarios.
`New customers get $300 in free credits to run, test, and deploy workloads.
`
`2. In the Google Cloud console, on the project selector page, select or create a Google Cloud project.
`
`Y Note: If you don't plan to keep the resources that you create in this procedure, create a project instead of selecting an
`existing project. After you finish these steps, you can delete the project, removing all resources associated with the
`project.
`
`Go to project selector
`
`3. Make sure that billing is enabled for your Google Cloud project.
`
`Source: https://cloud.google.com/video-intelligence/docs/annotate-video-command-line
`
`47. The Accused Product also practices the step of “converting audio files associated
`with the video to text.” For example, Google explains that Video Intelligence extracts audio data
`
`and converts that data into text using the “SPEECH_TRANSCRIPTION” feature. '’
`
`Cloud Video Intelligence API > Documentation > Guides Was this helpful? []b Qfl
`Speech transcription 0 - Send feedback
`
`Speech Transcription transcribes spoken audio in a video or video segment into text and returns blocks of text for each
`portion of the transcribed audio.
`
`Supported models
`
`The Video Intelligence only supports English (US). For other languages, use the Speech-to-Text API, which supports all
`available languages. For the list of available languages, see Language support in the Speech-to-Text documentation.
`
`To transcribe speech from a video, call the annotate method and specify SPEECH_TRANSCRIPTION inthe fea