DCT

1:19-cv-01088

Dynamic Data Tech LLC v. Dell Inc

Log In
Key Events
Amended Complaint

I. Executive Summary and Procedural Information

  • Parties & Counsel:
  • Case Identification: 1:18-cv-10454, S.D.N.Y., 02/04/2019
  • Venue Allegations: Plaintiff alleges venue is proper in the Southern District of New York because Defendant Dell transacts business, has committed acts of infringement, and maintains regular and established places of business in the district, including an office at One Penn Plaza in New York City.
  • Core Dispute: Plaintiff alleges that a wide range of Defendant’s computer products that incorporate H.265/High Efficiency Video Coding (HEVC) and VP9 video processing functionality infringe fourteen patents related to video compression, motion estimation, and image processing.
  • Technical Context: The technologies at issue involve fundamental techniques for video compression, which are critical for efficiently storing and transmitting high-definition digital video for applications such as streaming, video conferencing, and digital media playback.
  • Key Procedural History: The patents-in-suit originated from Philips, which the complaint describes as a pioneer in video and audio processing technologies. The complaint is a First Amended Complaint. Plaintiff notes that the patents-in-suit have been collectively cited as relevant prior art in over 400 U.S. and international patents by numerous major technology companies, which may be presented as evidence of industry significance.

Case Timeline

Date Event
1999-08-22 Priority Date for U.S. Patent No. 6,996,177
2000-05-18 Priority Date for U.S. Patent No. 7,010,039
2000-06-30 Priority Date for U.S. Patent No. 6,774,918
2001-09-12 Priority Date for U.S. Patent No. 7,929,609
2001-10-26 Filing Date for U.S. Patent No. 7,750,979 (no priority date listed on patent)
2002-01-17 Priority Date for U.S. Patent No. 8,073,054
2002-03-11 Priority Date for U.S. Patent No. 7,571,450
2002-12-19 Priority Date for U.S. Patent No. 8,135,073
2003-01-23 Priority Date for U.S. Patent No. 7,519,230
2003-04-03 Priority Date for U.S. Patent No. 7,542,041
2004-08-10 Issue Date for U.S. Patent No. 6,774,918
2005-06-03 Priority Date for U.S. Patent No. 7,894,529
2005-08-17 Priority Date for U.S. Patent No. 8,184,689
2006-02-07 Issue Date for U.S. Patent No. 6,996,177
2006-03-14 Issue Date for U.S. Patent No. 7,010,039
2006-10-17 Priority Date for U.S. Patent No. 8,189,105
2008-12-31 Filing Date for U.S. Patent No. 8,311,112 (no priority date listed on patent)
2009-04-14 Issue Date for U.S. Patent No. 7,519,230
2009-06-02 Issue Date for U.S. Patent No. 7,542,041
2009-09-01 Issue Date for U.S. Patent No. 7,571,450
2010-07-06 Issue Date for U.S. Patent No. 7,750,979
2011-02-22 Issue Date for U.S. Patent No. 7,894,529
2011-04-19 Issue Date for U.S. Patent No. 7,929,609
2011-12-06 Issue Date for U.S. Patent No. 8,073,054
2012-03-13 Issue Date for U.S. Patent No. 8,135,073
2012-05-22 Issue Date for U.S. Patent No. 8,184,689
2012-05-29 Issue Date for U.S. Patent No. 8,189,105
2012-11-13 Issue Date for U.S. Patent No. 8,311,112
2019-02-04 Complaint Filing Date

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 8,189,105 - *“Systems and Methods of Motion and Edge Adaptive Processing Including Motion Compensation Features”*

The Invention Explained

  • Problem Addressed: The complaint does not specify the problem addressed from the patent’s background section but describes the invention in the context of processing pixel information. The technical context of video deinterlacing implies a need to create high-quality progressive-scan frames from interlaced fields, which requires accurately estimating pixel values for missing lines, especially in the presence of motion and sharp edges where artifacts can occur (Compl. ¶ 21).
  • The Patented Solution: The invention discloses a system for processing pixel information using both motion and edge data (Compl. ¶ 21). A key component is a "blending component" that interpolates the intensity of a pixel. This interpolation is adaptive; its output depends on "motion reliability data." If the motion data is below a certain reliability threshold, the output is set to a first intensity estimate, suggesting a fallback to a more stable method when motion estimation is unreliable (’105 Patent, col. 4:1-12; Compl. ¶ 22). The system may also use segmentation to average contiguous pixels into a single segment during motion detection (Compl. ¶ 23).
  • Technical Importance: This approach of adaptively blending motion-compensated data with edge-based data aims to improve deinterlacing quality by reducing artifacts that occur when relying on a single method that may not be reliable in all scenarios (e.g., high motion, complex textures).

Key Claims at a Glance

  • The complaint asserts at least Claim 1 (Compl. ¶ 145).
  • The essential elements of independent claim 1 include:
    • A method for processing pixel information;
    • processing edge data from an edge-adaptive interpolation process, the edge data including a first intensity estimate for the pixel;
    • processing motion data from a motion compensation process, the motion data including first and second estimated motion vectors for a pixel in reference fields prior and subsequent to the present field;
    • determining a second intensity estimate for the pixel as a function of the edge data and the motion data; and
    • performing a blending process wherein final edge/motion data is calculated as a function of the first intensity estimate, the second intensity estimate, and motion reliability data.
  • The complaint does not explicitly reserve the right to assert dependent claims for this patent.

U.S. Patent No. 7,929,609 - *“Motion Estimation And/Or Compensation”*

The Invention Explained

  • Problem Addressed: The complaint states the invention improves various video signal processing functionalities, including motion compensated prediction for encoding, noise reduction, format conversion, and de-interlacing (Compl. ¶ 27). This suggests a need for more accurate and efficient motion estimation to reduce artifacts and improve compression in these applications.
  • The Patented Solution: The patent describes a method for estimating motion by defining an asymmetric search area around a selected image segment (’609 Patent, Abstract; Compl. ¶ 29). The center of this search area is intentionally offset from the center of the image segment being analyzed. This offset, which defines the search area's asymmetry, is statistically determined from an average vector of motion vectors established for previous images (Compl. ¶ 31). This allows the search to be biased in a direction of expected motion, potentially improving efficiency and accuracy.
  • Technical Importance: By using an asymmetric search area based on prior motion, the system can focus its computational resources where motion is most likely to be found, which can lead to faster and more accurate motion vector determination compared to a simple, symmetric search.

Key Claims at a Glance

  • The complaint asserts at least Claim 1 (Compl. ¶ 191).
  • The essential elements of independent claim 1 include:
    • A method of estimating or compensating motion in video images;
    • selecting an image segment of a given video image;
    • defining an asymmetric search area surrounding the image segment based on ranges of possible motion vectors; and
    • retrieving image data related to the asymmetric search area.
  • The complaint does not explicitly reserve the right to assert dependent claims for this patent.

U.S. Patent No. 8,135,073 - *“Enhancing Video Images Depending On Prior Image Enhancements”*

  • Technology Synopsis: The patent discloses enhancing subsequent images in a video stream where frames are predictively encoded based on previous frames (Compl. ¶ 34). It involves determining a re-mapping strategy for video enhancement on a decoded first frame and applying that strategy to both the first frame and corresponding regions of a second frame, thereby reducing processing requirements (Compl. ¶¶ 35, 40).
  • Asserted Claims: At least Claim 14 (Compl. ¶ 221).
  • Accused Features: The decoding functionality in Dell products that are compliant with the H.265/HEVC standard (Compl. ¶¶ 210, 221).

U.S. Patent No. 8,073,054 - *“Unit For And Method Of Estimating A Current Motion Vector”*

  • Technology Synopsis: The patent describes a method for motion estimation that generates a set of candidate motion vectors from previously estimated vectors (Compl. ¶ 45). To improve convergence speed, it adds a further candidate motion vector to the set, which is calculated based on a first and a second motion vector from the set of previously estimated vectors (Compl. ¶¶ 44, 47).
  • Asserted Claims: At least Claim 1 (Compl. ¶ 266).
  • Accused Features: The motion vector estimation functionality in Dell products compliant with the H.265/HEVC standard (Compl. ¶¶ 257, 266).

U.S. Patent No. 6,774,918 - *“Video Overlay Processor with Reduced Memory And Bus Performance Requirements”*

  • Technology Synopsis: The patent claims a method for providing an overlay (e.g., a cursor) on a display by downloading on-screen display (OSD) data in segments separated by gaps (Compl. ¶¶ 52, 54). During these gaps, a portion of the overlay data is downloaded, reducing the on-chip memory required to store the entire overlay (Compl. ¶¶ 55, 57).
  • Asserted Claims: At least Claim 18 (Compl. ¶ 296).
  • Accused Features: The video processing functionality in Dell products compliant with the H.265/HEVC standard, which allegedly segments data into Network Abstraction Layer (NAL) units that can include video (VCL) and overlay (non-VCL) data (Compl. ¶¶ 291, 297, 299).

U.S. Patent No. 8,184,689 - *“Method Video Encoding And Decoding Preserving Cache Localities”*

  • Technology Synopsis: The patent discloses reducing processing time and power consumption in video encoding/decoding by minimizing off-chip memory accesses (Compl. ¶ 62). The method involves providing a subset of image data from a second memory (e.g., main memory) into a first memory (e.g., cache) and simultaneously encoding/decoding more than one image by sharing access to the data subset (Compl. ¶¶ 64, 65).
  • Asserted Claims: At least Claim 1 (Compl. ¶ 336).
  • Accused Features: The encoding and decoding functionality of integrated graphics processors in Dell products, which allegedly use on-chip caches (first memory) and system DRAM (second memory) to process video frames (Compl. ¶¶ 317, 323, 325).

U.S. Patent No. 6,996,177 - *“Motion Estimation”*

  • Technology Synopsis: The patent describes a motion estimation method that involves a block-based process to find a "most frequently occurring block-based motion vector" (Compl. ¶¶ 69, 70). This most frequent vector is then used in a global motion vector estimation process to obtain a global motion vector, which is then, in turn, applied as a new candidate vector back into the block-based estimation process, creating a feedback loop (Compl. ¶¶ 71, 72).
  • Asserted Claims: At least Claim 1 (Compl. ¶ 353).
  • Accused Features: The motion estimation functionality in Dell products compliant with the H.265/HEVC standard (Compl. ¶¶ 345, 353).

U.S. Patent No. 7,010,039 - *“Motion Estimator for Reduced Halos in MC Up-Conversion”*

  • Technology Synopsis: The patent addresses the problem of poor motion estimation in areas where objects cover or uncover the background (Compl. ¶ 76). The solution involves using a criterion function that optimizes candidate vectors by performing the optimization at the temporal position of the next image in covering areas and at the temporal position of the previous image in uncovering areas (Compl. ¶ 80).
  • Asserted Claims: At least Claim 13 (Compl. ¶ 396).
  • Accused Features: The motion detection and video processing functionality in Dell products compliant with the H.265/HEVC standard (Compl. ¶¶ 384, 396).

U.S. Patent No. 8,311,112 - *“System And Method For Video Compression Using Predictive Coding”*

  • Technology Synopsis: The patent discloses a video compression system that performs predictive coding on a macroblock of a video frame by splitting the task (Compl. ¶ 84). An intra-frame coding unit codes a first set of pixels using reference pixels from the same frame, while an inter-frame coding unit codes the rest of the macroblock using reference pixels from at least one other video frame (Compl. ¶¶ 85, 86).
  • Asserted Claims: At least Claim 11 (Compl. ¶ 434).
  • Accused Features: The video compression functionality in Dell products compliant with the H.265/HEVC standard (Compl. ¶¶ 415, 434).

U.S. Patent No. 7,894,529 - *“Method And Device For Determining Motion Vectors”*

  • Technology Synopsis: The patent claims a method for determining motion vectors for individual image regions (Compl. ¶ 91). An image is subdivided into blocks, and a motion vector is assigned to a first image block. The method then determines a second image block through which the first block's motion vector passes, generates a modified motion vector as a function of the second block's vector, and assigns this modified vector back to the first block (Compl. ¶¶ 92-95).
  • Asserted Claims: At least Claim 1 (Compl. ¶ 458).
  • Accused Features: The motion vector determination functionality in Dell products compliant with the H.265/HEVC standard (Compl. ¶¶ 452, 458).

U.S. Patent No. 7,519,230 - *“Background Motion Vector Detection”*

  • Technology Synopsis: The patent discloses a method to select a correct background motion vector for a pixel in an occlusion region, which helps reduce "halos" in video images (Compl. ¶¶ 98, 99). The method involves computing a model-based motion vector for the pixel and comparing it with a set of existing motion vectors to select and assign the best fit as the background motion vector (Compl. ¶¶ 100-102).
  • Asserted Claims: At least Claim 6 (Compl. ¶ 498).
  • Accused Features: The video processing functionality in Dell products that contain VP9 encoding, which uses "segmentation" to identify background and foreground areas (Compl. ¶¶ 489, 496).

U.S. Patent No. 7,542,041 - *“Runtime Configurable Virtual Video Pipeline”*

  • Technology Synopsis: The patent describes a dynamically configurable multi-pipeline processing system (Compl. ¶ 105). The system includes multiple pipelines, each with a set of core pipeline elements, and a separate pool of auxiliary functional elements. These auxiliary elements can be selectively coupled into multiple pipelines as needed in response to external signals (Compl. ¶¶ 110-112).
  • Asserted Claims: At least Claim 1 (Compl. ¶ 514).
  • Accused Features: Dell products containing discrete NVIDIA GPUs, which allegedly comprise multiple vector pipelines (Graphics Processing Clusters) and a crossbar that allows for dynamic work migration between them (Compl. ¶¶ 511, p. 177).

U.S. Patent No. 7,571,450 - *“System For And Method Of Displaying Information”*

  • Technology Synopsis: The patent discloses a method for displaying information that allows user selections made in a first service to persist when switching to a second service (Compl. ¶ 117). A user selection of a type of information is used to filter and select a data element from the first service; when switching, a second step of filtering selects a semantically related data element from the second service based on the original user selection (Compl. ¶¶ 122, 124).
  • Asserted Claims: At least Claim 8 (Compl. ¶ 527).
  • Accused Features: The video processing functionality in Dell products compliant with the H.265/HEVC standard, which allegedly processes transport streams containing video and data elements (services) (Compl. ¶¶ 523, 527).

U.S. Patent No. 7,750,979 - *“Pixel-Data Line Buffer Approach Having Variable Sampling Patterns”*

  • Technology Synopsis: The patent discloses a method for motion compensation that uses decoupled line buffers to deliver a fixed number of pixels to a video processing stage (Compl. ¶ 128). It uses a variable window size for sampling, where the window size is a multiple of a fixed-size sampling-window (Compl. ¶¶ 131, 134). Pixels are stored in a first set of line buffers and then prefetched into a second set before being fetched by the processing stage (Compl. ¶¶ 135, 136).
  • Asserted Claims: At least Claim 1 (Compl. ¶ 571).
  • Accused Features: Dell products with Intel integrated graphics processors, which allegedly use memory controllers and caches that function as line buffers for video signal processing (Compl. ¶¶ 553, 558, 562).

III. The Accused Instrumentality

Product Identification

  • The complaint names a vast array of Dell computer products, including desktops, laptops, and all-in-one devices from product lines such as Inspiron, Latitude, Optiplex, Precision, Vostro, XPS, and Alienware (Compl. ¶¶ 140, 174, 210, etc.). Collectively, these are referred to as the "Dell '105 Product(s)," "Dell '609 Product(s)," etc., for each asserted patent.

Functionality and Market Context

  • The core accused functionality is the products' implementation of video processing compliant with the H.265/HEVC video compression standard (Compl. ¶ 141). The complaint alleges this functionality is provided by graphics processing units (GPUs) from NVIDIA and Intel that are incorporated into the Dell products (Compl. pp. 29-30). Plaintiff provides excerpts from Dell technical manuals and marketing materials stating that the products' video engines decode H.265/HEVC in hardware, which forms a central basis for the infringement allegations (Compl. ¶ 143, p. 31). The complaint also accuses products with VP9 encoding functionality of infringing certain patents (Compl. ¶ 489).

IV. Analysis of Infringement Allegations

U.S. Patent No. 8,189,105 Infringement Allegations

Claim Element (from Independent Claim 1) Alleged Infringing Functionality Complaint Citation Patent Citation
a method for processing pixel information, comprising: processing edge data from an edge-adaptive interpolation process, wherein the edge data includes a first intensity estimate for the pixel... Dell's HEVC-compliant products perform merge mode estimation, which generates merge estimation regions (MERs) comprised of prediction units (PUs) that identify edge information and contain luma values (intensity estimates). ¶¶ 148, 150-152 col. 4:1-5
processing motion data associated with motion compensation processing, wherein the motion data includes a first estimated motion vector for a pixel in a reference field prior to the present field... The accused products generate bi-directional prediction units containing two motion vectors that reference prior and subsequent frames, referred to as "reference picture list0" and "reference picture list1". ¶¶ 154-155 col. 4:13-19
and a second estimated motion vector for a pixel in a reference field subsequent to the present field; This functionality is alleged to be part of the bi-directional prediction process described in the HEVC standard. ¶¶ 154-155 col. 4:13-19
determining a second intensity estimate for the pixel as a function of the edge data and the motion data; The accused products' bi-prediction functionality allegedly generates a second intensity estimate based on motion data (bi-directional motion vectors) and edge data (luma location and prediction block information). ¶¶ 156, 158, 158 col. 4:20-22
and performing a blending process wherein final edge/motion data of the pixel is calculated as a function of the first intensity estimate, the second intensity estimate, and motion reliability data... The accused products allegedly perform a mixing process using bi-prediction with average and weighted mixing, where reliability values (predFlagL0, predFlagL1) are used to generate prediction utilization and reliability. ¶¶ 159, 161 col. 4:23-29
  • Identified Points of Contention:
    • Scope Questions: The analysis will likely focus on whether the terminology of the HEVC standard maps directly to the claim terms. For example, a question may be whether the HEVC standard's "merge mode estimation" and use of "luma" values (Compl. ¶¶ 148, 152) falls within the scope of the patent's term "edge-adaptive interpolation process" that includes a "first intensity estimate."
    • Technical Questions: A key question will be whether the HEVC standard's "weighted prediction" process, which uses "predFlagL0" and "predFlagL1" values (Compl. ¶ 161), performs the specific function of the claimed "blending process," which must calculate final data as a function of first and second intensity estimates and "motion reliability data." The complaint provides a diagram showing how prediction units are used to generate edge data. This visual, from a technical paper on HEVC, is used to illustrate how the accused products' HEVC encoders create edge data from partitioned video frames (Compl. ¶ 148, p. 34).

U.S. Patent No. 7,929,609 Infringement Allegations

Claim Element (from Independent Claim 1) Alleged Infringing Functionality Complaint Citation Patent Citation
A method of estimating or compensating motion in video images, said method comprising: using a video processor to select an image segment of a given video image; Dell's HEVC-compliant products select a "coding unit" which comprises a selected image segment. ¶¶ 180, 195 col. 2:27-31
using the video processor to define an asymmetric search area surrounding the image segment based on ranges of possible motion vectors for the image segment; The HEVC standard implemented by the accused products uses eight different partition modes, four of which are asymmetric. The complaint alleges these asymmetric partitions constitute the claimed "asymmetric search area" and that a flag (amp_enabled_flag) enables their use. ¶¶ 188, 189, 196 col. 2:32-35
and using the video processor to retrieve image data related to the asymmetric search area. The accused products allegedly retrieve image data related to the motion search area, including a motion vector index that is sent from the encoder to the decoder. ¶¶ 184, 185, 197 col. 2:36-37
  • Identified Points of Contention:
    • Scope Questions: A central dispute may be whether an HEVC "partition mode" (Compl. ¶ 188) constitutes a "search area" as required by the claim. The defense may argue that a partition defines the structure of a coding unit, whereas a search area defines a region in a reference frame to look for matching blocks. The complaint presents a figure illustrating symmetric and asymmetric partitions. This visual is used to support the allegation that the accused products' use of asymmetric partitions in the HEVC standard meets the "asymmetric search area" limitation of the claim (Compl. ¶ 189, p. 55).
    • Technical Questions: The complaint alleges the asymmetric search area's offset is "statistically determines from an average vector of motion vectors established for one or more previous images" (Compl. ¶ 198), a limitation from a dependent claim. The basis for this in the accused HEVC standard functionality is alleged to be the selection of a temporal candidate that defaults to a position "just outside of the collocated prediction unit" (Compl. ¶ 187). It raises the question of whether this default positioning scheme meets the "statistically determined" limitation.

V. Key Claim Terms for Construction

Patent No. 8,189,105

  • The Term: "blending process"
  • Context and Importance: This term is critical because Claim 1 requires the final pixel data to be calculated as a function of two different intensity estimates and "motion reliability data." Plaintiff maps this to the HEVC standard's weighted prediction process (Compl. ¶¶ 159, 161). Practitioners may focus on whether the HEVC standard's process, which uses prediction flags, performs the specific three-input calculation required by the claim.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The specification describes the blending component in functional terms as performing "an intensity-calculation procedure" without being limited to a single mathematical formula, which may support construing the term to cover various mixing or weighting schemes (’105 Patent, col. 4:23-25).
    • Evidence for a Narrower Interpretation: The detailed description of an embodiment specifies that the blending component "interpolates the intensity of the pixel... to equal to the first intensity estimate if motion reliability data is below a threshold" (’105 Patent, col. 4:8-12). This could support a narrower construction limited to a threshold-based switch rather than a continuous weighting.

Patent No. 7,929,609

  • The Term: "asymmetric search area"
  • Context and Importance: The concept of an "asymmetric search area" appears central to the patent's novelty. Plaintiff alleges that the "asymmetric partitions" used in the HEVC standard meet this limitation (Compl. ¶¶ 188, 196). Practitioners may focus on whether a "partition" of a coding unit is structurally and functionally equivalent to a "search area" for motion vectors in a reference frame.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The claims do not appear to further define "search area," which could support giving the term its plain and ordinary meaning in the art, potentially encompassing any asymmetric region used in motion estimation. The abstract broadly refers to defining "an asymmetric search area surrounding the image segment" (’609 Patent, Abstract).
    • Evidence for a Narrower Interpretation: The specification's description of the prior art discusses the search area as a "surrounding sub-area" in a preceding image used for searching for a matching pixel block (’609 Patent, col. 1:40-45). This could support a narrower definition tied to a region in a reference frame, which a defendant might argue is distinct from a partition of a current coding unit.

VI. Other Allegations

  • Indirect Infringement: The complaint alleges inducement of infringement for all asserted patents. The allegations are based on Dell providing products with the accused HEVC functionality and further providing "documentation and training materials that cause customers and end users" to use the products in an infringing manner (e.g., Compl. ¶¶ 166-168, 203-204). The complaint cites to specific Dell user manuals and setup guides as evidence of such inducement (e.g., Compl. fn. 46, fn. 72).
  • Willful Infringement: For each patent, the complaint alleges that Dell has had knowledge of the patent "since at least service of this First Amended Complaint or shortly thereafter" (e.g., Compl. ¶ 167). It further alleges Dell's infringement is "willful, wanton, malicious, in bad faith, deliberate, consciously wrongful, flagrant, or characteristic of a pirate" (e.g., Compl. ¶ 169). The allegations are based on post-suit knowledge.

VII. Analyst’s Conclusion: Key Questions for the Case

  • A core issue will be one of standardized infringement: will compliance with the mandatory portions of the H.265/HEVC standard be found to necessarily practice the methods claimed in the patents-in-suit? This will require the court to first construe the claims and then perform a detailed technical comparison of the construed claims against the standard's specifications.
  • The case will also turn on a question of terminological and functional mapping: can patent-specific terms like "asymmetric search area" (’609 Patent) and "blending process" (’105 Patent) be construed to encompass the distinct technical features and terminology of the HEVC standard, such as "asymmetric partitions" and "weighted prediction"? The outcome will depend on whether the court finds a fundamental match or a mismatch in technical operation and definition.