US 12088843

Here is a concise summary of US patent 12088843:

Title: Method for deriving a motion vector

Assignee: Velos Media LLC (Current Assignee and Original Assignee)

Inventors: Christopher A. Segall

Filing Date: 2023-09-21

Issue Date: 2024-09-10

Abstract: A method for decoding video includes creating a first list of motion vectors from at least one neighboring block in a current frame of the video and creating a second list of motion vectors from at least one previous block in a temporally earlier frame of the video. A third list of motion vectors is created based upon the first list and the second list. Based upon receiving a motion vector competition control parameter selecting one of the motion vectors from the third list, wherein the second list of motion vectors is further based upon a flooring function.

Plain-Language Overview of Independent Claims:

• Claim 1 (Method Claim): This claim describes a method for predicting motion vectors in a video bitstream. It involves three main steps:

  1. Determining a co-located block's location: This is done based on the current block's coordinates and includes performing an arithmetic right shift operation.
  2. Deriving a motion vector candidate: A co-located block motion vector candidate is derived from the location determined in the previous step.
  3. Selecting a motion vector: A final motion vector is chosen from a group of candidates that includes the co-located block motion vector candidate.

• Claim 9 (Non-transitory Computer-Readable Storage Medium Claim): This claim covers a non-transitory computer-readable storage medium. This medium stores instructions that, when executed by processing circuitry, perform the exact same operations as described in independent Claim 1 for motion vector prediction.

• Claim 13 (Device Claim): This claim describes a device designed for motion vector prediction in a video bitstream. The device includes a processor and a memory. The memory contains instructions that, when executed by the processor, cause the device to perform the exact same operations as described in independent Claim 1.

CAFC 2026 Dockets:
No specific dockets or cases for US patent 12088843 were found in the CAFC 2026 dockets from the search results.

As a patent attorney, I have investigated known litigation involving US patent 12088843.

Based on my current search, no specific litigation cases for US patent 12088843 were found in the CAFC 2026 dockets or on Unified Patents' litigation portal. While PACER (Public Access to Court Electronic Records) is a comprehensive source for federal court records, specific case numbers are generally required for direct searches within PACER. General searches for patent numbers on PACER or the CAFC website without specific case information are not feasible due to the design of those systems. Therefore, I cannot definitively state there is no litigation without direct case numbers.

However, the Google Patents page for US12088843B2 indicates that "Family has litigation" and points to two US cases:

• A US case filed in California Central District Court, case number 2:26-cv-03052.
• A US case filed in Texas Western District Court, case number 1:25-cv-00967.

As of today, April 26, 2026, and based on the information provided, the outcomes or current statuses of these specific cases are not detailed. To obtain that information, a direct search within the respective court's electronic filing system (likely PACER, if a subscription is held) using the provided case numbers would be necessary.

Proceedings overview

There is no PTAB activity on file for US patent 12088843.

Strategic summary

As of 2026-07-22, there are no AIA trial proceedings (Inter Partes Review, Post-Grant Review, or Covered Business Method) on file for US patent 12088843. This means all claims (1-20) of US12088843B2 remain untested by the PTAB.

The absence of PTAB activity suggests that the patent has not yet faced validity challenges in this forum. This could be due to several reasons, such as the patent being relatively new (granted 2024-09-10), not yet widely asserted, or the claims being considered robust against common prior art challenges.

Recommended next steps

If facing an assertion of US patent 12088843, a potential defendant has a clear path to file an AIA trial petition without being estopped by prior PTAB proceedings. Given the patent's recent grant date, a Post-Grant Review (PGR) might be an option if filed within nine months of the grant, allowing for challenges under § 101, § 102, § 103, and § 112. After this nine-month window, or at any time, an Inter Partes Review (IPR) could be filed, challenging claims under § 102 and § 103 based on patents and printed publications. The absence of prior PTAB activity means there are no existing final written decisions to analyze for claim construction or validity determinations that would bind future proceedings or appeals.

Inventors

• Christopher A. Segall: Employer at the time of filing is not determinable from the provided patent text.

Original assignee

The original assignee on the issued patent US12088843B2 is Velos Media LLC. Information about whether Velos Media LLC shipped a product embodying the claims, their primary line of business, or their current operational status (operating, acquired, dissolved, in bankruptcy) is not explicitly stated in the provided patent text or its abstract.

Assignment timeline

No recorded assignments for US patent 12088843B2 were found on the USPTO Assignment Center. The patent is currently assigned to Velos Media LLC as the original assignee.

Timeline diagram

timeline
    title Ownership of US 12088843
    2011 : Priority date
    2023 : Filed by Velos Media LLC
    2024 : Issued to Velos Media LLC

NPE / troll-pattern signals

1. Shell-entity transfer — unclear. While Velos Media LLC's business model is centered on patent licensing, and it was reportedly founded by Marconi Group to create new licensing platforms for video coding technologies, the provided information does not contain explicit evidence of a shell entity transfer for this specific patent (e.g., no intermediate transfer to an entity with "IP / Patents / Licensing / Holdings / Ventures" in its name, nor a registered-agent address listed for the company itself within the patent record). Velos Media was acquired by TechStream (New York) on January 1, 2025. Velos Media describes its business as a "provider of licensing solutions for video codec patents" and has "hundreds of fundamental HEVC patents throughout the world."

2. Known asserter in the chain — present. Velos Media LLC is identified as a high-frequency plaintiff and former Avanci licensor, having filed patent infringement suits, including against ByteDance (TikTok), asserting patents broadly directed to video coding, including US12088843. Unified Patents has also challenged patents owned by Velos Media. RPX Corporation, a provider of patent risk solutions, has published articles mentioning Velos Media's litigation activity.

3. Repeat correspondent across the chain — not present. No assignments were found on the USPTO Assignment Center, thus no correspondent information is available to assess recurrence.

4. Cascading transfers — not present. No assignments were found on the USPTO Assignment Center.

5. Pre-litigation transfer — unclear. The patent was issued to Velos Media LLC, and the first litigation was filed against ByteDance (TikTok) by Velos Media on June 22, 2025. Since Velos Media is the original assignee, there isn't a transfer event within 6 months of litigation for this patent. However, Velos Media's business model is explicitly patent licensing and assertion.

6. Bankruptcy fire-sale — not present. No evidence of bankruptcy proceedings for the original assignee, Velos Media LLC, was found.

7. Privateering — unclear. While Velos Media acquires patents from operating companies (e.g., Sharp, BlackBerry, Panasonic, Qualcomm, Sony) to license them, the provided information doesn't explicitly state that Velos Media asserts these patents on behalf of the original assignors against their competitors in a privateering arrangement. However, the model is consistent with an operating company transferring patents to an NPE.

8. Defensive aggregator (anti-NPE) — not present. The chain does not terminate at a known defensive aggregator. Instead, Velos Media LLC is actively asserting its patents.

Verdict

NPE — high confidence

The verdict is high confidence NPE due to the current assignee, Velos Media LLC, being a known patent licensor and asserter, actively engaged in patent infringement litigation for this specific patent. Velos Media's business model centers on licensing portfolios of standard-essential patents from other companies. This aligns with the profile of a non-practicing entity.

USPTO Assignment Center search: https://assignmentcenter.uspto.gov/patent/index.html (search for patent number US12088843).

To identify the most relevant prior art for US patent 12088843, I will examine the "Patent Citations" section of the patent itself. The USPTO Public Search tool can be used to confirm patent details.

Based on the information provided in the patent text, the following are the patent citations listed under "Citations":

Patent Citations for US12088843B2:

1. US20090074069A1 (Byeong Moon Jeon)

   • Full Citation: US20090074069A1 - Method of deriving a motion vector of a bi-predictive block based on a motion vector of a co-located block in a reference picture.
   • Publication Date: 2009-03-19.
   • Filing Date: 2002-10-04.
   • Brief Description: This patent application describes a method for deriving a motion vector of a bi-predictive block based on a motion vector of a co-located block in a reference picture.
   • Potential Anticipated Claim(s) under 35 U.S.C. § 102: This reference is highly relevant to Claim 1, Claim 9, and Claim 13, as it describes a method that involves "deriving a motion vector... based on a motion vector of a co-located block," which directly relates to the concept of deriving a co-located block motion vector candidate as taught in these independent claims. The general concept of using co-located block motion vectors for prediction is disclosed.

2. US20110122950A1 (Ji Tianying)

   • Full Citation: US20110122950A1 - Video decoder and method for motion compensation for out-of-boundary pixels.
   • Publication Date: 2011-05-26.
   • Filing Date: 2009-11-26.
   • Brief Description: This patent application describes a video decoder and method for motion compensation, specifically addressing out-of-boundary pixels.
   • Potential Anticipated Claim(s) under 35 U.S.C. § 102: While it relates to video decoding and motion compensation, its primary focus on "out-of-boundary pixels" might not directly anticipate the specific steps of "determining a location corresponding to a co-located block" using an arithmetic right shift, or the selection from a plurality of candidates based on a control parameter, as precisely defined in claims 1, 9, and 13. However, elements relating to motion compensation in a decoder context could be broadly relevant to the overall system described in the claims.

3. US20120128060A1 (Mediatek Inc.)

   • Full Citation: US20120128060A1 - Method and Apparatus of Spatial Motion Vector Prediction.
   • Publication Date: 2012-05-24.
   • Filing Date: 2010-11-23.
   • Brief Description: This patent application describes a method and apparatus for spatial motion vector prediction.
   • Potential Anticipated Claim(s) under 35 U.S.C. § 102: This reference is highly relevant to Claim 1, Claim 9, and Claim 13 due to its focus on "spatial motion vector prediction." The claims of US12088843B2 involve deriving motion vector candidates from neighboring blocks (spatial) and a co-located block (which can be considered a form of spatial/temporal prediction). Depending on the specifics of how "spatial motion vector prediction" is performed in US20120128060A1, it could potentially anticipate the steps of determining a co-located block's location and selecting a motion vector from candidates. Further analysis would be needed to determine if the arithmetic shift operation and specific candidate selection mechanism are disclosed.

It is important to note that the detailed analysis for anticipation would require a thorough claim construction and comparison of each claim element to the disclosures in these prior art documents. The brief descriptions above provide an initial assessment of potential relevance.

Obviousness Analysis under 35 U.S.C. § 103 for US12088843

This analysis identifies combinations of prior art references that would render the claims of US patent 12088843 obvious to a person having ordinary skill in the art (POSITA), along with the motivation to combine them. A POSITA in this field would likely be an engineer or computer scientist working on video compression and decoding, familiar with standards like H.264/AVC and HEVC.

For an invention to be obvious, there must be a reason for a skilled artisan to combine the elements of prior art in the way the new invention does, with a reasonable expectation of success. Mere knowledge of individual claim elements in the prior art is not sufficient to establish a motivation to combine.

Independent Claims 1, 9, and 13

Independent Claims 1 (method), 9 (non-transitory computer-readable storage medium), and 13 (device) all share the same core limitations:

1. Determining a co-located block's location: This involves using current block coordinates and an arithmetic right shift operation.
2. Deriving a co-located block motion vector candidate: This candidate is based on the determined co-located block location.
3. Selecting a motion vector: The selection is made from a plurality of candidates, including the co-located block motion vector candidate.

Combination 1: US2009/0074069A1 in view of US2012/0128060A1 (or Bossen et al.)

• US2009/0074069A1 (Jeon): This patent application, titled "Method of deriving a motion vector of a bi-predictive block based on a motion vector of a co-located block in a reference picture," explicitly discloses deriving a motion vector for a current block based on a motion vector of a co-located block in a reference picture. Jeon teaches selecting a motion vector of the co-located block in a reference frame for direct mode motion vector calculation. This directly addresses elements 2 and 3 of the independent claims, as it involves deriving a co-located block motion vector candidate and using it in a selection process (implicitly, as it's used for direct mode calculation).
• US2012/0128060A1 (Mediatek Inc.) / US8824558B2: This patent document, titled "Method and Apparatus of Spatial Motion Vector Prediction," discusses improvements in motion vector prediction, including the use of spatial and temporal MVP candidates. It describes deriving motion vector predictor candidates from neighboring prediction unit partitions, as well as temporal motion vector prediction. Crucially, the detailed description of US12088843 states: "A modified technique for the current block in the current frame that spans the pixel locations (x,y) to (x+N, y+M) is to select and to store in the memory buffer the motion vector from the previous frame used for the pixel prediction located at ((x>>Z)<<Z, (y>>Z)<<Z) where Z is an integer." This describes a flooring function using arithmetic right and left shift operations. While US2009/0074069A1 focuses on the concept of using a co-located block motion vector, it doesn't explicitly detail the method of determining the co-located block's precise location using a bitwise shift operation. US2012/0128060A1 (and its issued counterpart US8824558B2) details various motion vector prediction techniques, including Advanced Motion Vector Prediction (AMVP) in HEVC, which uses spatial and temporal candidates. The description of how motion vector predictors are derived from neighboring prediction units (subclause 8.5.3.1.6) and temporal luma motion vector prediction (subclause 8.5.3.1.7) indicates the kind of detailed coordinate and block processing that would make the arithmetic shift operation a known technique for location determination. Further, the context of motion vector prediction in video coding often involves manipulating block coordinates efficiently, and bitwise shift operations are a standard method for division and multiplication by powers of two, often used for block alignment or downsampling.
• Motivation to Combine: A POSITA would be motivated to combine Jeon's concept of using co-located block motion vectors with the efficient coordinate manipulation techniques (like bitwise shifts for flooring functions) common in video coding, as described or implied by US2012/0128060A1. The goal of motion vector prediction is to reduce computational complexity and memory requirements while maintaining high compression efficiency. Using a "flooring function" (implemented with arithmetic right and then left shifts) to determine the co-located block's location, as suggested in US12088843, directly contributes to memory buffer reduction by selecting fewer candidate vectors, which is a known problem in the art. This combination would improve the efficiency of motion vector prediction by providing a more precise and computationally lightweight method for identifying relevant co-located motion vectors, leading to better compression performance or reduced processing load. This is a "known-technique" rationale for combining, where the bit shift operation from US2012/0128060A1 (or general knowledge in the field) is applied to Jeon's co-located motion vector prediction to achieve a known improvement in efficiency.

Combination 2: US2009/0074069A1 in view of general knowledge in the art (e.g., H.264/AVC, H.263, MPEG-2 standards)

• US2009/0074069A1 (Jeon): As discussed above, Jeon discloses the derivation of a motion vector for a bi-predictive block based on a motion vector of a co-located block in a reference picture.
• General Knowledge in Video Coding: Video compression standards like H.264/AVC, H.263, and MPEG-2 extensively utilize motion estimation and compensation. These standards divide images into blocks (e.g., macroblocks of 16x16 pixels) and predict motion vectors from neighboring or co-located blocks. The "flooring function" implemented by arithmetic right and left shifts is a fundamental and well-known bitwise operation used for efficient calculation of block coordinates or downsampling in digital signal processing, particularly in systems dealing with image and video data. For instance, (x >> Z) << Z is a common way to find the nearest multiple of 2^Z less than or equal to x, effectively "flooring" x to a grid. Given that video coding heavily relies on block-based processing and efficient arithmetic, a POSITA would be aware of and routinely employ such bitwise operations for manipulating block coordinates.
• Motivation to Combine: The motivation to combine Jeon's teaching with this general knowledge stems from the continuous drive in video coding to enhance efficiency and reduce computational overhead. By applying a standard, efficient bitwise operation (arithmetic right shift for coordinate determination) to the process of identifying co-located blocks as described by Jeon, a POSITA would achieve a more optimized method for deriving motion vector candidates. This directly addresses the stated problem in US12088843 of reducing memory requirements and computational complexity in motion vector prediction. The use of bitwise shifts for block alignment or quantization of coordinates is an obvious design choice for optimizing performance in block-based video processing.

Dependent Claims

Many of the dependent claims elaborate on aspects already addressed by the independent claims or represent common implementations in video coding.

• Claim 2, 10, 14 (Neighboring Block Candidates): These claims add deriving at least one neighboring block motion vector candidate. This is explicitly taught in US2012/0128060A1 (AMVP uses spatial MVP candidates from neighboring groups of the current block) and is a standard practice in motion vector prediction (e.g., MVB from above, MVC to the left, MVD to the above left).
• Claim 3, 11, 15 (Top-Left Coordinates and N/2, M/2 offsets): These claims specify using top-left coordinates and adding half the horizontal/vertical size (N/2, M/2) before applying the shifting operation. US12088843 itself describes this as a "modified technique" for selecting motion vectors, noting that (x+N/2, y+M/2) is a common center-point reference. The combination of using a center point for a block and then applying a flooring function (via bit shifts) to derive a co-located reference point would be an obvious design choice for a POSITA seeking to reduce candidate list size while still referencing a central, representative part of the block in the previous frame.
• Claim 4, 12, 16 (Arithmetic Left Shift): These claims specify performing an arithmetic left shift on the result of the right shift with the same magnitude. As explained in US12088843, this pair of operations (>>Z then <<Z) is "generally equivalent to 'the floor of (x/(2^Z))'*(2^Z)." This is a well-known and standard programming idiom for implementing a flooring function to the nearest multiple of 2^Z, and its application in conjunction with coordinate determination would be obvious for efficiency.
• Claim 5, 17 (Magnitude of 4): These claims specify the magnitude Z as 4. The choice of the magnitude for the shift operation (Z) would be a design parameter chosen by a POSITA based on desired block size granularity (e.g., 4x4 or 16x16 blocks mentioned in US12088843).
• Claim 6, 18 (Co-located block in previous frame): These claims specify the co-located block being in a previous frame. This is fundamental to temporal motion vector prediction and explicitly taught by US2009/0074069A1 and is a core concept in video compression generally.
• Claim 7, 19 (Neighboring Block Locations): These claims specify specific neighboring block locations (above, above left, left). These are standard locations for deriving spatial motion vector candidates in various video coding standards (e.g., H.264/AVC, HEVC).
• Claim 8, 20 (Encoding or Decoding): These claims specify encoding or decoding the current block based on the selected motion vector. This is the ultimate purpose of motion vector prediction in video compression and is inherently obvious once a motion vector has been derived.

In summary, the core inventive step of US12088843 lies in applying an efficient "flooring function" (via arithmetic bit shifts) to determine the location of a co-located block for motion vector prediction. This specific implementation of coordinate determination, when combined with the known concept of using co-located motion vectors for prediction (as taught by Jeon), and motivated by the continuous need for efficiency and memory reduction in video coding, would be obvious to a POSITA.