PTAB

IPR2025-01345

Snap Inc v. Nokia Technologies Oy

Log In
Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Method and Apparatus for Providing Complexity Balanced Entropy Coding
  • Brief Description: The ’701 patent discloses a method for video coding that categorizes syntax elements into two groups based on their frequency of occurrence relative to a threshold. Less frequently occurring elements are entropy coded using a more complex process that involves context updates, while more frequently occurring elements are entropy coded using a simpler process that bypasses context updates to reduce computational complexity and improve efficiency on devices with limited resources.

3. Grounds for Unpatentability

Ground 1: Claims 1, 6, 8, 13, 15, and 20 are obvious over Murashita

  • Prior Art Relied Upon: Murashita (Patent 5,650,783).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Murashita, which relates to adaptive arithmetic coding for general data, discloses a system that categorizes symbols based on their frequency of occurrence against a predetermined threshold. Murashita’s system performs a context updating process for symbols with a cumulative frequency below the threshold (a first category) and bypasses this updating process for symbols with a cumulative frequency above the threshold (a second category). Petitioner contended that Murashita’s “symbols” are equivalent to the claimed “syntax elements” and that its threshold-based decision to update or bypass context maps directly to the limitations of the independent claims. The use of a cumulative frequency was argued to be an obvious proxy that still relies on the underlying frequency of occurrence, thus meeting the claim limitation.
    • Motivation to Combine (for §103 grounds): While a single-reference ground, Petitioner asserted it would have been an obvious modification for a person of ordinary skill in the art (POSITA) to apply Murashita’s data coding method to video content, a common and known data type. It would also be an obvious simplification to base the comparison on individual symbol frequency instead of cumulative frequency to reduce calculations and increase processing speed, a goal explicitly stated in Murashita.
    • Expectation of Success: A POSITA would have a high expectation of success in applying Murashita's established data compression techniques to video, a well-known application for such methods.

Ground 2: Claims 1-20 are obvious over Murashita in view of Marpe

  • Prior Art Relied Upon: Murashita (Patent 5,650,783) and Marpe (Application # US 2005/0253740).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner asserted that Murashita provides the high-level method of categorizing syntax elements based on frequency to determine whether context updating is performed. Marpe, which discloses a well-known Context-Adaptive Binary Arithmetic Coding (CABAC) engine for video compression, provides the specific, concrete implementation for the two types of entropy coding. Marpe’s Figure 2 depicts a standard CABAC architecture that explicitly includes a binarization stage, a "regular" coding path with a context modeler (for context updates), and a "bypass" coding path that skips the context modeler. A switch directs data to one path or the other. The combination was argued to teach every limitation of the independent claims. Further, Marpe's explicit disclosure of a binarization stage and its discussion of processing video in "slices" were argued to render dependent claims 3 and 2 obvious, respectively.
    • Motivation to Combine: A POSITA would combine these references to achieve efficiency gains. Marpe itself explains that its binary arithmetic coding (CABAC) engine is more efficient and computationally less complex than the multi-level arithmetic coding used by Murashita. A POSITA would be motivated to replace Murashita's older coding engine with Marpe's more efficient and well-known CABAC engine, using Murashita's frequency-based threshold logic to control the existing switch between Marpe’s regular and bypass coding paths.
    • Expectation of Success: A POSITA would have a high expectation of success, as the combination involves integrating two known, compatible video coding concepts. Murashita’s control signal (a high or low signal based on frequency comparison) could directly and predictably control Marpe's existing architectural switch.

Ground 3: Claims 1, 3, 6-8, 10, 13-15, 17, and 20 are obvious over Yu

  • Prior Art Relied Upon: Yu (Wei Yu & Yun He, A High Performance CABAC Decoding Architecture, 51 IEEE 1352 (Nov. 2005)).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Yu, a technical paper on a high-performance CABAC decoding architecture, renders the challenged claims obvious. Yu explicitly classifies syntax elements into two classes based on their frequency of occurrence: a first class of low-frequency elements ("occur only once or a few times in one MB") and a second class of high-frequency elements ("occur tens or hundreds of times in one MB"). Yu teaches that high-occurrence elements are processed in a "bypass mode" where "no context model is needed," thereby bypassing context updates. Conversely, low-occurrence elements are processed in a "regular mode" that utilizes a corresponding context model, thereby being subjected to context updates. This was argued to map directly to the claimed categorization and differential processing.
    • Motivation to Combine (for §103 grounds): This is a single-reference ground. Petitioner argued that a POSITA would have found it obvious to apply the principles from Yu's decoder to a corresponding encoder, as encoding and decoding processes within video standards like H.264/AVC are well-understood, complementary operations.
    • Expectation of Success: The expectation of success would be high because the underlying logic of using frequency to select a processing path (bypass vs. regular) is identical for both encoding and decoding, making the application of Yu’s teachings to an encoder a predictable design choice.

4. Relief Requested

  • Petitioner requests institution of an inter partes review and cancellation of claims 1-20 of Patent 9,036,701 as unpatentable.