Google LLC v. Uniloc 2017 LLC

1. Case Identification

• Patent #: 6,349,154
• Filed: January 28, 2020
• Petitioner(s): Google LLC
• Patent Owner(s): Uniloc 2017 LLC
• Challenged Claims: 1-4

2. Patent Overview

• Title: Method of Creating a High-Resolution Picture
• Brief Description: The ’154 patent discloses a method for creating a high-resolution still picture from a sequence of lower-resolution pictures. The method involves estimating motion between the lower-resolution pictures with sub-pixel accuracy and using this estimated motion to construct the final high-resolution image.

3. Grounds for Unpatentability

Ground 1: Claims 1-4 are obvious over Hwang in view of Yoon.

• Prior Art Relied Upon: Hwang (Patent 5,666,160) and Yoon (Patent 5,532,747).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Hwang disclosed the core elements of claim 1: a method for creating a high-resolution image by receiving a sequence of lower-resolution pictures, estimating motion between them with sub-pixel accuracy, and using that motion to generate an enhanced-resolution output image. Hwang’s system (Fig. 6) detailed a motion detector, an image interpolator, and an image synthesizer to achieve this. However, Petitioner contended that Hwang did not explicitly disclose the specific technique for motion detection. The petition asserted that the limitation of "subjecting the sequence of pictures to motion-compensated predictive encoding, thereby generating motion vectors representing motion between successive pictures" was not taught by Hwang.
  • Motivation to Combine (for §103 grounds): Petitioner asserted a person of ordinary skill in the art (POSITA) would have been motivated to combine Hwang's high-resolution imaging system with Yoon's well-known motion-compensated predictive encoding technique. Yoon described a standard, MPEG-based hybrid coder for compressing video signals for transmission, a technique Petitioner argued was the "most effective" and widely used at the time. The motivation stemmed from the need to transmit video data over limited-bandwidth channels, a common requirement in applications like surveillance or remote sensing where Hwang's technology would be useful. To make Hwang's system practical for remote applications, a POSITA would have naturally turned to a standard compression method like that in Yoon to reduce the data size for efficient transmission before applying Hwang's resolution-enhancement process at the receiver.
  • Expectation of Success (for §103 grounds): A POSITA would have had a reasonable expectation of success because combining the systems was a predictable application of known techniques. Yoon provided a standard solution (MPEG encoding) for a known problem (bandwidth limitation) that was fully compatible with the video processing taught by Hwang. The petition argued that integrating Yoon's encoder/decoder would not detract from Hwang's core functionality and would predictably result in a system capable of remote, high-resolution imaging.
  • Key Aspects: The petition included a "Demonstrative A" diagram illustrating how Yoon's MPEG-compliant encoder/decoder would be integrated into Hwang's system architecture. This diagram showed Yoon's encoder compressing the video at a remote location and the decoder decompressing it at the receiver before it is fed into Hwang's resolution-enhancement components.

4. Key Claim Construction Positions

• Petitioner contended that while no special constructions were necessary, the challenged claims were unpatentable even under the constructions adopted in a related district court case (Uniloc 2017 LLC v. Google LLC, Case No. 2:18-cv-00496 E.D. Tex.). Petitioner explicitly adopted these constructions for its analysis to demonstrate the strength of its invalidity arguments.
• "motion-compensated predictive encoding" (claims 1-4): Construed by the district court as "predictive encoding based on motion between a current picture and a previously encoded picture." Petitioner argued that Yoon’s MPEG-based system met this construction because it generates motion vectors by differentially combining a current picture with a previously encoded one.
• "recursively adding, in the high-resolution domain..." (claim 4): Construed by the district court as "recursively adding, in the high-resolution domain, the pixel values of a current decoded picture to the pixel values of a previously created picture..." Petitioner argued the combination met this, as Hwang’s image synthesizer (65) used an IIR filter that recursively added a current picture with a previously generated and stored picture (q(i;n-1)) on a pixel-by-pixel basis.

5. Relief Requested

• Petitioner requested institution of an IPR and cancellation of claims 1-4 of the ’154 patent as unpatentable.