PTAB

IPR2015-00343

Google Inc v. Network 1 Technologies Inc

Log In
Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Algorithms for Automatic Content Recognition
  • Brief Description: The ’179 patent relates to a computer-implemented method for identifying an unknown electronic media work and associating a corresponding action with it. The method involves maintaining a database of "compact electronic representations" of reference works, obtaining a similar representation from an unknown work, and identifying the unknown work using a "non-exhaustive neighbor search" to find a match in the database.

3. Grounds for Unpatentability

Ground 1: Obviousness over Levy in view of Arya - Claims 1-3, 6-14, 18-19, 21-26, and 30-37 are obvious over Levy in view of Arya.

  • Prior Art Relied Upon: Levy (Patent 6,505,160) and Arya (a 1998 publication titled "An Optimal Algorithm for Approximate Nearest Neighbor Searching in Fixed Dimensions").
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Levy disclosed all elements of the independent claims except for the "non-exhaustive neighbor search." Levy taught a system for identifying a media work by extracting an identifier or "fingerprint" from it, using that identifier to look up a corresponding action in a server database, and then executing the action. Arya disclosed an efficient, non-exhaustive algorithm for performing an "approximate nearest neighbor" search, which is a search for close but not necessarily exact matches.
    • Motivation to Combine: Petitioner contended that a person of ordinary skill in the art (POSITA) would have understood that fingerprinting algorithms, like those in Levy, could generate slightly different fingerprints from different sources of the same media work. This reality would necessitate searching for near matches, not just exact ones. A POSITA would combine Levy’s system with Arya’s well-known search algorithm to improve the robustness and accuracy of the identification process in real-world conditions.
    • Expectation of Success: Arya explicitly stated that nearest neighbor searching is a critical problem in applications including multimedia databases. A POSITA would therefore have had a high expectation of success in applying Arya’s established search method to the fingerprint database taught by Levy.

Ground 2: Anticipation by Conwell - Claims 1-3, 6, 8-14, 19, 21-26, 30-31, and 34-37 are anticipated by Conwell.

  • Prior Art Relied Upon: Conwell (Patent 6,970,886).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner asserted that Conwell disclosed every limitation of the challenged claims. Conwell taught a method for directing users of digital content to related websites by generating an identifier from the content (e.g., via a hashing algorithm), storing these identifiers with associated URLs in a database, and looking up the URL when a user plays the content. Petitioner argued that Conwell’s use of a hash table lookup constituted a "non-exhaustive neighbor search." The search was non-exhaustive because it did not require a brute-force comparison of all possible hashes. It was a "neighbor" search because Conwell’s hashing algorithm was designed to map similar, non-identical inputs to the same hash output, thereby identifying "neighbor" content.

Ground 3: Obviousness over Ghias in view of Philyaw - Claims 1-3, 8, 10-14, 18-19, 21-27, 29, 31, and 34-37 are obvious over Ghias in view of Philyaw.

  • Prior Art Relied Upon: Ghias (Patent 5,874,686) and Philyaw (Patent 6,098,106).

  • Core Argument for this Ground:

    • Prior Art Mapping: Petitioner argued that Ghias disclosed a system for identifying a melody hummed by a user. This system extracted a "contour representation" (a compact representation) of the hummed tune and searched a melody database using an efficient, approximate (i.e., non-exhaustive neighbor) search to find a match. Philyaw disclosed a system that linked broadcast content to an action, such as displaying an advertisement, by extracting an embedded identifier from the content and using it to query a server for a corresponding URL.
    • Motivation to Combine: A POSITA would combine Ghias’s content-based identification method with Philyaw’s system for linking content to actions. Philyaw’s reliance on pre-embedded identifiers was time-consuming and expensive. Ghias provided a superior method of identifying content based on its intrinsic characteristics. A POSITA would have been motivated to replace Philyaw's inefficient identification step with Ghias's more advanced fingerprinting and search method to create a system that could associate actions (like advertisements) with content without requiring embedded data.

  • Additional Grounds: Petitioner asserted an additional anticipation challenge based on Wood (Patent 7,743,092), which disclosed a query-by-humming system that calculated a distance score to identify songs and allowed for subsequent purchase.

4. Key Claim Construction Positions

  • "compact electronic representation" / "extracted features": Petitioner proposed these terms be construed together to mean "a quantitative representation of the features of a media work, or a subset of the features of a media work." This construction was argued to be consistent with the specification, which used the terms synonymously and described deriving a compact representation via feature extraction.
  • "non-exhaustive neighbor search": Petitioner proposed this term be construed as "a search that locates a match without conducting a brute force comparison of all possible matches, and all data within all possible matches." This construction was central to mapping prior art like Conwell, where a single hash table lookup was argued to be non-exhaustive, and Arya, which explicitly taught improvements over brute-force searches.
  • "neighbor search": Consistent with the above, this term was proposed to mean "identifying a close, but not necessarily exact, match." This was critical for the obviousness arguments, which relied on the need to find near matches for slightly different content fingerprints.

5. Relief Requested

  • Petitioner requested institution of an inter partes review and cancellation of claims 1-3, 6-14, 18-19, 21-27, and 29-37 of the ’179 patent as unpatentable.