PTAB

IPR2025-01263

Snap Inc v. Nokia Technologies Oy

Log In
Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Motion Vector Coding
  • Brief Description: The ’714 patent discloses methods for improving the efficiency of video compression standards like H.264/H.265. The technology focuses on reducing computational overhead by optimizing the process of removing redundant candidates from a motion vector prediction list, which is used to predict the movement of pixel blocks between frames. Specifically, it teaches comparing a potential motion vector candidate against a determined subset of other candidates, rather than performing an exhaustive comparison of every possible pair.

3. Grounds for Unpatentability

Ground 1: Obviousness over Rusert and Karczewicz - Claims 1-3, 5-10, 12-17, 19-24, 26-30 are obvious over Rusert in view of Karczewicz.

  • Prior Art Relied Upon: Rusert (Application # 2011/0194609) and Karczewicz (Application # 2011/0249721).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Rusert teaches the core elements of the ’714 patent’s method for video encoding and decoding. Rusert discloses a technique for improving motion vector prediction by creating a candidate list (PMV_CANDS) from a subset of previously coded motion vectors. Critically, Rusert teaches removing redundant or "unnecessary" candidates from this list by comparing a new candidate against those already added to the list. Because the list is built sequentially, this comparison is inherently against a subset of all available candidates, thus avoiding a comparison of every possible pair. Karczewicz was presented as providing the necessary context for the then-emerging H.265 video compression standard, including its terminology like Prediction Units (PUs), which are analogous to the "blocks" described in Rusert's H.264-based disclosure.
    • Motivation to Combine: Petitioner asserted that a person of ordinary skill in the art (POSITA) would have been motivated to combine the references. Rusert expressly stated that its efficiency-improving principles, though described in the context of H.264, could be applied to other coding standards. Karczewicz taught the structure of H.265, the direct successor to H.264. A POSITA would combine Rusert's known method for efficient candidate list de-duplication with the emerging H.265 framework taught by Karczewicz to achieve predictable improvements in coding efficiency. This combination represented a simple substitution of Karczewicz’s PUs for Rusert’s blocks.
    • Expectation of Success: A POSITA would have had a high expectation of success. The combination involved applying a known technique (Rusert's candidate reduction) to a similar, next-generation system (H.265 as described by Karczewicz). As both references address complementary aspects of block-based video encoding, the result—improved compression efficiency—was entirely predictable.

Ground 2: Obviousness over Rusert, Karczewicz, and Lin - Claims 1-30 are obvious over Rusert and Karczewicz in view of Lin.

  • Prior Art Relied Upon: Rusert (Application # 2011/0194609), Karczewicz (Application # 2011/0249721), and Lin (Application # 2014/0092981).
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground builds upon the combination of Rusert and Karczewicz to further address limitations related to blocks divided into multiple prediction units (e.g., claim 4). Petitioner argued that Lin explicitly teaches a method for removing redundant motion vector candidates in scenarios where a coding unit is divided into a first PU and a second PU. Lin teaches that because the two PUs in a divided block often have different motion, the motion vector from one PU is not a good predictor for the other and can be excluded from its candidate list. This directly addresses the claimed steps of examining whether a block is divided and excluding a candidate from the list if it belongs to the second prediction unit.
    • Motivation to Combine: A POSITA, having already combined Rusert and Karczewicz, would be motivated to incorporate Lin's teachings to further optimize the candidate reduction process. Lin provides a straightforward solution for a common and known scenario in video coding—divided blocks. Applying Lin's logic would further Rusert's primary goal of "reduc[ing] the number of previous motion vectors that must be considered," thereby increasing coding efficiency in a well-understood manner.
    • Expectation of Success: The combination would yield predictable results. Adding Lin's specific technique for handling divided blocks to the general framework of Rusert and Karczewicz is a simple application of a known principle to improve a known process. The outcome would be a more robust candidate reduction algorithm with enhanced efficiency, which was the expected goal.

4. Key Claim Construction Positions

  • "the block": Petitioner adopted the construction from prior related IPRs, defining this term as "the block associated with the first spatial motion vector prediction candidate." This construction is central to their argument that the subset of candidates for comparison is determined relative to the candidate currently being evaluated.
  • "a subset of … candidates": Petitioner argued this phrase means a subset comprising one or more candidates. This construction is important because it supports their assertion that Rusert's method of comparing a new candidate to the existing members of a progressively built list (which could contain only one member early in the process) satisfies the claim limitation.

5. Relief Requested

  • Petitioner requests institution of an inter partes review and cancellation of claims 1-30 of the ’714 patent as unpatentable.