Unified Patents, LLC v. Uniloc 2017 LLC

1. Case Identification

• Case #: IPR2019-01126
• Patent #: 6,519,005
• Filed: May 28, 2019
• Petitioner(s): Unified Patents Inc.
• Patent Owner(s): Uniloc 2017 LLC
• Challenged Claims: 1-16 and 39-42

2. Patent Overview

• Title: Method and Apparatus for Performing Motion Estimation for Multiple Prediction Modes
• Brief Description: The ’005 patent describes a method for performing motion estimation in digital video compression, such as for the MPEG-2 standard. The patent asserts as novel a scheme for concurrently searching for an optimal macroblock match according to multiple different motion prediction modes during a single search operation to determine the optimum mode, rather than selecting a mode beforehand.

3. Grounds for Unpatentability

Ground 1: Claims 1-16 and 39-42 are obvious over Ishihara and Applicant's Admitted Prior Art (AAPA).

• Prior Art Relied Upon: Ishihara (“A Half-pel Precision MPEG2 Motion-Estimation Processor with Concurrent Three-Vector Search,” a 1995 IEEE publication) and Applicant’s Admitted Prior Art (AAPA) from the ’005 patent specification.
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Ishihara discloses a hardware processor for MPEG-2 motion estimation that meets the core limitations of the challenged claims. Ishihara’s processor concurrently calculates distortion metrics (Mean Absolute Difference, or MAD) for multiple prediction modes (e.g., 16x16 frame prediction, 16x8 field prediction) using a specialized summation circuit. A "minimum-value detector" then analyzes these concurrently generated distortion metrics to select the "best vectors." Petitioner asserted this process is equivalent to the ’005 patent’s claimed method of concurrently performing motion estimation for a plurality of modes to determine an "optimum" mode.
  • Motivation to Combine: The combination with AAPA was primarily for context and to address specific dependent claims. For claim 6, which requires providing picture-type information (e.g., I-, P-, B-type), Petitioner argued that AAPA establishes this as a fundamental part of the MPEG standard. A person of ordinary skill in the art (POSITA) would be motivated to use this standard picture-type information in Ishihara’s MPEG-2 compliant system to enable proper motion compensation (e.g., bidirectional prediction for B-type pictures), which Ishihara supports.
  • Expectation of Success: A POSITA would have a high expectation of success because Ishihara already taught the core concurrent processing, and applying standard MPEG-2 features described in the AAPA, such as picture types, was a routine implementation detail for achieving compatibility and full functionality.

Ground 2: Claims 1-16 and 39-42 are obvious over Mombers, Nakajima, and Senda.

• Prior Art Relied Upon: Mombers (“A Video Signal Processor Core for Motion Estimation in MPEG2 Encoding,” a 1997 IEEE publication), Nakajima (Patent 5,412,435), and Senda (“A Simplified Motion Estimation Using An Approximation for the MPEG-2 Real-Time Encoder,” a 1995 conference paper).

• Core Argument for this Ground:

  • Prior Art Mapping: Petitioner contended that Mombers discloses a video signal processor that evaluates costs for all MPEG-2 prediction types. For each candidate motion vector, Mombers’ pixel processor accumulates matching costs for both the entire macroblock (16x16) and its halves (16x8), making it possible to evaluate multiple prediction modes concurrently. However, Petitioner noted Mombers does not explicitly teach selecting the mode with the smallest error. To supply this, Petitioner introduced Nakajima, which teaches a motion estimation system with a comparator that explicitly selects the mode providing the "smallest prediction error signal." For the limitation of "generating a motion vector in response" to the determination, Petitioner introduced Senda, which teaches a half-pel refinement search performed after an initial full-pel search (of the type described in Mombers) to generate a more accurate final motion vector.
  • Motivation to Combine: A POSITA would combine Mombers with Nakajima to improve coding efficiency by implementing a known, simple technique—selecting the mode with the minimum error—to achieve the predictable result of an optimal prediction. The motivation to further combine this with Senda was to improve the accuracy of the output vectors. Using a half-pel search to refine the result of a full-pel search was a known method to improve motion estimation and reduce artifacts, and a POSITA would have predictably applied Senda's refinement technique to Mombers' system.
  • Expectation of Success: A POSITA would have a reasonable expectation of success because the combination involved applying known techniques (minimum error selection from Nakajima, half-pel refinement from Senda) to improve a similar system (Mombers' MPEG-2 processor) to achieve predictable benefits (optimality and accuracy).

• Additional Grounds: Petitioner asserted that claims 39-40 are obvious over Mombers and Nakajima, and that claim 6 is obvious over Mombers, Nakajima, Senda, and AAPA. These grounds relied on the same core logic and prior art teachings detailed in the grounds above.

4. Key Claim Construction Positions

• “Optimum” Prediction Mode: Petitioner argued that, based on the ’005 patent specification, this term should be construed to mean "the mode that yields the smallest value of the error metric." This construction was central to Petitioner's argument that selecting the "best" mode based on the lowest error, as taught by the prior art (e.g., Nakajima), directly maps to the claimed "optimum" mode.
• “Best” Match: Similarly, Petitioner argued that the specification defines the "best match" as the pixel array "having the smallest value of the corresponding sum [of errors]." This supported the argument that prior art systems determining a match by minimizing an error metric (like MAD) were finding the "best match" as recited in the claims.

5. Arguments Regarding Discretionary Denial

• Petitioner argued that discretionary denial under §325(d) would be inappropriate. While Nakajima was cited in an Information Disclosure Statement (IDS) during the original prosecution of the ’005 patent, Petitioner contended that the Examiner never substantively addressed it. Therefore, the arguments and combinations presented in the petition, which rely heavily on Nakajima's teachings, were not previously considered by the USPTO.

6. Relief Requested

• Petitioner requested the institution of an inter partes review and the cancellation of claims 1-16 and 39-42 of the ’005 patent as unpatentable.