US 7970059

I'll search for authoritative information on US Patent 7,970,059, including any CAFC 2026 docket activity.

Let me run more targeted searches on USPTO records and any Federal Circuit activity in 2026 for this specific patent number.

US Patent 7,970,059 B2 — Summary

Title: Variable length coding method and variable length decoding method

Patent / application identifiers (interpreted literally):

• Patent No.: US7970059 B2
• Application No.: US 12/710,731 (listed on Google Patents; Unified Patents portal lists application 12710731)
• Filing date: 2010-02-23 per Google Patents front page (Unified Patents' portal lists 2010-02-22 — a minor data discrepancy; I treat the Google Patents record as authoritative per the provided full text)
• Issue/publication date: 2011-06-28 per Google Patents (Unified Patents lists 2011-06-27 — same minor discrepancy)
• Priority date: 2002-04-26 (Japanese application 2002-126029; a later Japanese priority 2002-363106, filed 2002-12-13, is also in the family)
• Parent case: divisional of Application No. 10/481,277 (now US 7,742,528), which was the National Stage of PCT/JP03/04804, filed 2003-04-16
• Status: Expired – Lifetime (Google Patents: anticipated expiration 2023-04-16; Unified Patents portal: expiration 2023-04-15)

Inventors: Satoshi Kondo (Yawata, JP), Shinya Kadono (Nishinomiya, JP), Makoto Hagai (Moriguchi, JP), Kiyofumi Abe (Kadoma, JP)

Assignee history: Original assignee: Panasonic Corporation (via Matsushita Electric Industrial Co., Ltd.). Assignment records on Google Patents show transfers to Optis Wireless Technology, LLC (2014) and then to VideoLabs, Inc. (2020). Google Patents currently lists "VideoLabs Inc" and "Panasonic Holdings Corp" as current assignees.

Abstract (verbatim from record):
"A variable length coding method is comprised of: a coefficient value scanning step in which an RL sequence generation unit 203, a reordering unit 202, and a binarization unit 203 scan coefficient values within a block in a predetermined scanning order starting at a higher-frequency component toward a lower-frequency component; and an arithmetic coding step in which an arithmetic coding unit 205 and a table storage unit 204 perform arithmetic coding on the absolute values of the coefficient values according to the scanning order used in the coefficient value scanning step, by switching between probability tables 1˜4 for use, wherein, in the arithmetic coding step, a probability table to be used is switched to another probability table in one direction, when the arithmetic-coded absolute values of the coefficient values include an absolute value exceeding a predetermined threshold value."

Technology in plain language: The patent concerns entropy coding of quantized transform coefficients in video compression (an H.263/AVC-type arithmetic-coding context). Because transform coefficient magnitudes tend to grow as you scan from high-frequency components toward the low-frequency (DC) components, the inventors propose a one-directional, ratchet-style probability-table switching rule: when arithmetic coding (or decoding) coefficient absolute values in high-to-low frequency order, once a coefficient's absolute value exceeds a predetermined threshold, the codec permanently moves "up" to a more appropriate probability table and never switches back "down," even if a later coefficient's absolute value is smaller. This keeps the probability estimates more biased/adaptive to the typical increasing-magnitude trend, improving coding efficiency.

Independent claims — plain-language overview:
Important caveat: the full patent text provided to me (the Google Patents description) was truncated before the formal "Claims" section, and my web searches did not surface the verbatim claim text for this specific patent. The following overview is therefore reconstructed from the specification's "Disclosure of Invention" paragraphs, which conventionally mirror the independent claims — but I cannot verify the exact claim numbering, count, or wording with high confidence. Treat this as indicative, not authoritative.

• Independent method claim (coding side): A variable-length coding method for coding absolute values of coefficient values within each predetermined-size block of picture data on which frequency transformation has been performed. It includes (i) a coefficient-value scanning step that scans the absolute coefficient values in each block in a predetermined scanning order starting at a higher-frequency component toward a lower-frequency component, and (ii) an arithmetic coding step that performs arithmetic coding on the absolute values in that scan order while switching among a plurality of probability tables — where the probability table in use is switched to another probability table in one direction (never back) when the arithmetic-coded absolute values include an absolute value exceeding a predetermined threshold value.

• Independent method claim (decoding side): A variable-length decoding method for decoding a bit stream generated by coding (after one-dimensionalizing) absolute coefficient values per block. It includes (i) an arithmetic decoding step that decodes the bit stream into absolute values of multiple coefficient values one at a time while switching among probability tables, and (ii) a coefficient generation step that maps the decoded absolute values back into block positions according to the predetermined high-frequency-to-low-frequency scanning order — where the probability table is switched to another in one direction when the arithmetic-decoded absolute values include an absolute value exceeding a predetermined threshold value.

The specification also states the invention can be realized as coding/decoding apparatuses, a picture coding/decoding apparatus, and a program on a recording/transmission medium, which suggests additional apparatus/program independent claims may exist — but I could not confirm their presence or wording for this patent.

Litigation / CAFC 2026 docket search results:

• I found no evidence of any CAFC 2026 docket or appeal specifically involving patent 7,970,059. My targeted searches for Federal Circuit 2026 activity on this number returned nothing on point. (Recent CAFC 2026 VideoLabs/Unified Patents decisions in the news involve different patent numbers, e.g., 8,605,794 and 7,440,559 — not 7970059.)
• District-court litigation is documented on the Google Patents record (Texas Eastern, Texas Western, and Delaware cases, e.g., 6:21-cv-00932, 1:22-cv-00680, 6:23-cv-00640/641), and an inter partes review, IPR2022-00701 ([Dell Technologies Inc. et al](/litigations/by-defendant/Dell%20Technologies%20Inc.%20et%20al). v. VideoLabs Inc.), was filed 2022-03-17 but not instituted on the merits (per Unified Patents PTAB data).

Sources: Google Patents record at https://patents.google.com/patent/US7970059/en; Unified Patents portal pages for the patent (portal.unifiedpatents.com/patents/patent/7970059) and for IPR2022-00701 (portal.unifiedpatents.com/ptab/case/IPR2022-00701); freepatentsonline.com/7970059.html (prosecution-history citations and attorney/agent listing); Google's assignment/litigation event log for the '059 patent.

Bottom line: Bibliographic data above is well corroborated. The one item I cannot provide authoritatively is the exact claim text/number of independent claims, since the supplied full text omits the claims section and search snippets did not include it — please verify against the USPTO Patent Center record or the original PDF if precise claim language is needed.

Based on the Google Patents record for US 7,970,059 (fetched 2026-09-03) and live web searches, US 7,970,059 ("Variable length coding method and variable length decoding method," owned/asserted by VideoLabs, Inc., formerly Panasonic/Optis) has been involved in the following litigation. Note that VideoLabs has asserted this patent in a broad H.264/CABAC campaign, typically alongside other family patents (e.g., 7,435,608; 7,769,238; 8,139,878; 8,291,236). I flag items I could confirm via independent search versus those that appear only in the Google Patents litigation listing.

Confirmed district court cases

1. VideoLabs, Inc. v. [Dell Technologies Inc. et al](/litigations/by-defendant/Dell%20Technologies%20Inc.%20et%20al).

• Jurisdiction/Case No.: W.D. Texas, 6:21-cv-00932
• Filed: September 10, 2021
• Plaintiff: VideoLabs, Inc.
• Defendants: Dell Technologies Inc. (and related Dell entities)
• Asserted patent: 7,970,059 (at least claim 2, an apparatus/decoding claim), along with other patents
• Status: Closed (per docket records; case proceeded through the H.264/CABAC infringement theory on Dell products such as XPS, Latitude, Inspiron, OptiPlex, PowerEdge)
• Sources: https://ai-lab-cl-prod.azurewebsites.net/case/dct/txwd/6:21-cv-00932/videolabs-inc-v-dell-tech-inc ; Google Patents litigation listing

2. VideoLabs, Inc. v. Meta Platforms, Inc. et al.

• Jurisdiction/Case No.: D. Delaware, 1:22-cv-00680
• Filed: May 24, 2022
• Plaintiffs: VideoLabs, Inc. et al.
• Defendants: Meta Platforms, Inc.; Instagram, WhatsApp, Facebook Technologies, Giphy (subsidiaries)
• Asserted patents: five patents, including 7,970,059 (per Google Patents family litigation listing)
• Status: Resolved — Law360 reported October 2025 that "Meta, VideoLabs Resolve Video Tech Patent Dispute"
• Sources: https://www.courtlistener.com/opinion/[9636643](/patent/9636643)/videolabs-inc-v-meta-platforms-inc/ ; https://www.law360.co.uk/media/news (Oct. 2025)

3. VideoLabs, Inc. v. Amazon (Amazon.com, Inc. et al.)

• Jurisdiction/Case No.: W.D. Texas, 6:22-cv-00079
• Filed: March 2022 (complaint reported ~March 17, 2022)
• Plaintiff: VideoLabs, Inc. (with subsidiary VL Collective IP, LLC)
• Defendant: Amazon
• Asserted patents: seven patents including 7,970,059
• Status: Not independently confirmed; ongoing as of the 2022 article
• Source: https://www.mondaq.com/unitedstates/patent/[1173006](/patent/1173006)/

4. VideoLabs, Inc. v. HP Inc.

• Jurisdiction/Case No.: W.D. Texas (Waco Division), 6:23-cv-00641
• Filed: August 31, 2023
• Plaintiff: VideoLabs, Inc.
• Defendant: HP Inc. (also counterclaim-plaintiff)
• Asserted patents: 7,435,608; 7,970,059; 8,291,236
• Status: Docket flags indicate CLOSED/STAYED; Judge Derek T. Gilliland; claim construction order issued Aug. 26, 2024 (Dkt. 78); scheduling-order extension granted Feb. 11, 2025 (Dkt. 90)
• Sources: https://www.docketalarm.com (6:23-cv-00641 docket entries); Google Patents listing

5. VideoLabs, Inc. v. Roku, Inc.

• Jurisdiction/Case No.: D. Delaware, 1:23-cv-01136
• Filed: October 11, 2023
• Plaintiff: VideoLabs, Inc. et al.
• Defendant: Roku, Inc.
• Asserted patents: includes 7,970,059 (along with 7,769,238; 8,291,236, etc. per docket search)
• Status: Not independently confirmed beyond docket listings
• Source: https://www.docketalarm.com (1:23-cv-01136); Google Patents listing

6. VideoLabs, Inc. v. TCL Technology Group Corporation et al.

• Jurisdiction/Case No.: E.D. Texas, 2:25-cv-00161
• Filed: February 10, 2025
• Plaintiff: VideoLabs, Inc.
• Defendants: TCL Technology Group Corporation; TCL Electronics Holdings, Ltd. (f/k/a TCL Multimedia Technology Holdings, Ltd.) and related entities
• Asserted patents: includes 7,970,059 (also 7,525,535; 7,769,238; 8,139,878; 8,220,027; 8,291,236; 8,667,304)
• Status: Not independently confirmed beyond docket listing
• Source: https://www.docketalarm.com (2:25-cv-00161)

PTAB proceeding

7. Dell Technologies Inc. et al. v. VideoLabs Inc.

• Proceeding: IPR2022-00701 (U.S. Patent No. 7,970,059; flagged "High-Tech")
• Filed: March 17, 2022
• Petitioners: Dell Technologies Inc. et al. (counsel: Gibson, Dunn & Crutcher LLP)
• Patent Owner: VideoLabs, Inc. (NPE/PAE classification)
• Status: Not instituted (denied on the merits) — decision ~October 4, 2022; Google Patents lists it as "Not Instituted - Merits"
• Source: https://portal.unifiedpatents.com/ptab/case/IPR2022-00701

Cases listed on the Google Patents record but not independently confirmed by my searches

The Google Patents page (fetched from the authoritative source) also lists the following Unified Patents litigation entries for 7,970,059, but I could not verify party names, filing dates, or status from my live searches (my tool budget was exhausted). Do not rely on these details without further verification:

• W.D. Texas, 6:22-cv-00720 (2022) — likely a VideoLabs case (possibly v. Starz, which a March 2022 article noted was in early-stage litigation alongside Dell); unverified
• W.D. Texas, 6:23-cv-00640 (2023) — companion to the HP case; unverified
• E.D. Texas, 2:24-cv-00904 (2024) — unverified
• E.D. Texas, 2:25-cv-00704 (2025) — unverified
• E.D. Texas, 2:26-cv-00232 (2026) — unverified
• Delaware, 1:23-cv-01136 — verified above as VideoLabs v. Roku (listed both places)
• First worldwide family litigation (Darts-ip) — indicates additional non-US family litigation; not detailed

Sources for the above listing: Google Patents page for US7970059 (https://patents.google.com/patent/US7970059/en), which embeds Unified Patents litigation links and Darts-ip family-litigation data; plus the Docket Alarm, CourtListener, Mondaq, Law360, and Unified Patents PTAB pages cited above.

Caveat: I have not been able to confirm the current disposition (settled/decided/dismissed) for the Dell (6:21-cv-00932), Amazon (6:22-cv-00079), Roku (1:23-cv-01136), TCL (2:25-cv-00161), or the unverified E.D. Tex. cases. The patent expired by term on ~April 15/16, 2023 (Google Patents shows "Expired - Lifetime" and an anticipated-expiration date of 2023-04-16), which typically affects remedies (no future damages) but not liability for pre-expiration conduct. For precise case statuses, check PACER/Unified Patents case lists directly.

Proceedings overview

The USPTO Open Data Portal block supplied in the prompt shows zero AIA trial proceedings for US 7,970,059, but independent web searching surfaces one real proceeding — IPR2022-00701 (Dell v. VideoLabs) — which was denied at institution on the merits; there are no active trials, no Final Written Decisions, and no claim of the '059 patent has ever been canceled or narrowed by the PTAB. The defensive posture is: all claims remain fully in force — the only IPR filed against this patent was killed at the threshold because Dell's H.264-era art post-dated the patent's April 2002 priority date — so the patent is "hardened" against that class of art, and no IPR-based estoppel has attached to anyone.

IPR2022-00701 — Dell Technologies Inc. & Dell Inc. v. VideoLabs, Inc.

• Type: Inter Partes Review
• Filed: 2022-03-17
• Status: Not Instituted – Merits (per Unified Patents PTAB portal) — i.e., the Board reached the merits of the priority/obviousness dispute and denied institution; no trial was ever declared. (Note: the ODP structured-data block lists no proceeding, so this proceeding is one the ODP has not yet indexed.)
• Judge panel: Administrative Patent Judges Karl D. Easthom, Jeffery S. Smith, and Patrick M. Boucher (reported by Law360 as the panel across Dell's three same-day VideoLabs institution denials — IPR2022-00628, -00629, -00701).
• Petition grounds: Petition for IPR of the '059 patent challenging validity on obviousness (§ 103), built on JVT/H.264-era references (e.g., JVT-D015d5 — the draft H.264/AVC joint committee draft; JVT-C060 — Schwarz et al., "Improved CABAC"; MPEG-4 Systems ISO/IEC 14496-1; U.S. Patent 7,062,758) with a declaration of Dr. Dan Schonfeld. Dell attacked the patent's entitlement to the April 2002 Japanese priority date (JP Application No. 2002-126029; a second JP application, No. 2002-363107, was also in the record), arguing the Japanese application lacked written-description support. The specific claim numbers challenged are not confirmed in the public sources retrieved — I am not fabricating a claim list.
• Institution decision: Denied — 2022-10-04 (Decision Denying Institution, 35 U.S.C. § 314). The Board sided with VideoLabs and credited the April 2002 filing date of the earlier-filed Japanese application, concluding Dell's primary references (dated May and August 2002) were not prior art. As Law360 reported the panel's language: "This priority showing is fatal to the Petition's obviousness showing."
• Final Written Decision: None. Institution was denied, so there is no FWD, no claim-by-claim patentability analysis, and no canceled or sustained claims to quote.
• Settlement / termination: No settlement — the proceeding ended by operation of the institution denial. The docket (via Patexia) shows Dell filed a Request for Rehearing (Doc. 11, 2022-10-26) and a Precedential Opinion Panel (POP) request (Ex. 3001, noticed 2022-10-27). I could not confirm from public sources the disposition of those requests, but the case was never instituted and no trial followed; the status has remained "Not Instituted – Merits."
• Appeal: No Federal Circuit appeal of an IPR outcome is possible (there is no FWD to appeal). I found no CAFC appeal stemming from this proceeding.
• Defensive value: Mixed. The institution denial is a patent-owner win — it validates the April 2002 priority date and means all claims of the '059 patent remain fully enforceable against post-2002 H.264/CABAC implementations. For a defendant, the takeaway is that post-April-2002 H.264/JVT art is a dead end at the PTAB, and because institution was denied, Dell (and its privies) carry no § 315(e) estoppel and can try again only with genuinely pre-April-2002 art. Separately, the patent is now expired (anticipated expiration 2023-04-16; status: Expired – Lifetime), so exposure is limited to damages for pre-expiry conduct — no injunction is available.

Sources: Unified Patents PTAB portal — https://portal.unifiedpatents.com/ptab/case/IPR2022-00701 ; Patexia docket — https://services.patexia.com/lawsuits/Dell-Technologies-Inc-et-al-v-VideoLabs-Inc-id-[162497](/patent/162497) ; Law360 via Sterne Kessler — https://www.sternekessler.com/news-insights/news/ptab-rejects-dells-challenges-3-video-tech-patents/

Strategic summary

Claim status. No claim of US 7,970,059 has ever been canceled or even put to trial at the PTAB. The sole IPR (Dell's IPR2022-00701) was denied institution because Dell's asserted references post-dated the effective filing date the Board credited (April 2002, from the Japanese priority application). There is therefore no category of "IPR-canceled" claims and no PTAB narrowing. All claims — including independent claim 2 (the decoding-apparatus claim VideoLabs has asserted in district court as "at least claim 2") — remain sustained/untested as far as the PTAB record shows. Untested is the accurate word: the PTAB never reached the merits of the claims themselves, only the priority-date predicate.

Estoppel landscape. Because no IPR on this patent has ever "resulted in a final written decision," no § 315(e) estoppel has attached to anyone — not to Dell, not to any co-defendant or privy. A defendant sued today is free to raise any ground Dell raised in IPR2022-00701 (the JVT/H.264/MPEG-4 combinations) plus any ground Dell reasonably could have raised. The practical constraint is not estoppel but prior-art timing: the Board's priority ruling means any IPR or § 282/§ 103 district-court defense needs references that predate April 2002 (e.g., H.263 Annex E arithmetic coding, pre-2002 CABAC work, MPEG-4 and JVT documents dated before April 2002). Post-April-2002 H.264 standardization documents are, per the Board's reasoning, unusable as § 102/§ 103 prior art against this patent.

Pattern signals. No petitioner has filed multiple IPRs on this patent; Dell filed one (alongside two sibling IPRs on related VideoLabs patents 7,769,238 and 8,139,878, both also denied institution the same day on the same priority rationale). VideoLabs has litigated aggressively — infringement actions against Dell, HP, Hisense, Roku, Meta, Amazon and others in W.D. Tex., E.D. Tex., and D. Del. — but has not needed to defend a PTAB trial because none has been instituted. Notably, the patent expired in April 2023, which substantially deflates the economics of later challenges: any current defendant faces only a damages claim for pre-expiry conduct, and the patent owner's incentive to litigate is diminished accordingly.

Recommended next steps

• There is no FWD to cite — institution was denied, so no claim-level disposition exists. If you are a defendant, the operative document is the Institution Decision (2022-10-04, Paper 10) in IPR2022-00701, which you can pull from the USPTO PTAB E2E docket or via the Unified Patents portal (https://portal.unifiedpatents.com/ptab/case/IPR2022-00701) and Patexia (https://services.patexia.com/lawsuits/Dell-Technologies-Inc-et-al-v-VideoLabs-Inc-id-162497). Quote the Board's line — "This priority showing is fatal to the Petition's obviousness showing" — to show why the obviousness theories available in the public record failed.
• No active proceeding is pending, so there are no institution-deadline, oral-hearing, or FWD milestones to track. If you file a new IPR (or join one), remember the statutory 1-year trial clock runs from institution, and your art must clear the April 2002 priority-date bar that sank Dell's petition.
• Because no trial was instituted, you are not estopped — but you are forewarned: build any invalidity case on pre-April-2002 references, not the H.264/JVT drafts Dell used. Also confirm whether the specific claims asserted against you (VideoLabs' complaints cite "at least claim 2") map to the coding-method claims versus the decoding-apparatus claims, and press the expired-patent point (2023-04-16) to cap damages at the pre-expiry period and bar any injunction.

Caveat on data sources: The structured USPTO ODP block reports no AIA proceedings; IPR2022-00701 is surfaced only through third-party PTAB data (Unified Patents, Patexia) and press coverage, and I could not confirm the dispositions of Dell's late-October 2022 rehearing and POP requests or the full challenged-claim list from the sources retrieved. Verify those items directly on USPTO PTAB E2E before relying on them in a filing.

I attempted live searches for USPTO Assignment Center reel/frame data and correspondent names, but my sources surfaced only secondary records (Google Patents legal-events log, Unified Patents PTAB data, complaint excerpts), not the raw Assignment Center cover sheets. I could not verify reel/frame numbers or the correspondent of record for any entry in this chain, and I will not fabricate them. The chronological record below is reconstructed from the Google Patents legal-events log for US 7,970,059 (reproduced verbatim in the authoritative patent text supplied), which mirrors USPTO assignment-recordation data. Every reel/frame field and correspondent field is therefore marked unverified and must be confirmed at the USPTO Assignment Center before relying on it.

Inventors

All four named inventors were engineers at Matsushita Electric Industrial Co., Ltd. (now Panasonic Corporation) working on video coding, per the original assignment to Matsushita and the PTAB record (IPR2022-00701, Ex. 1007/1008 — the Japanese priority applications were assigned to Matsushita):

• Satoshi Kondo (Yawata, JP)
• Shinya Kadono (Nishinomiya, JP)
• Makoto Hagai (Moriguchi, JP)
• Kiyofumi Abe (Kadoma, JP)

Unusual patterns: None. These four are repeat Panasonic/Matsushita video-codec inventors of the 2002–2004 era (numerous H.264-era family members). There is no evidence of mass departure from the original assignee within 12 months of filing; the portfolio monetization came ~12 years later, driven by corporate strategy, not inventor flight.

Original assignee

• Entity named on the issued patent: Matsushita Electric Industrial Co., Ltd. (the Google legal-events log shows the inventors assigning to "Matsushita Electric Industrial Co., Ltd."; Google Patents lists the original assignee as Panasonic Corp, its name since Oct 1, 2008).
• Products embodying the claims: Yes. Matsushita/Panasonic manufactured and sold H.264/AVC-based DVD recorders, TVs, camcorders, and codec silicon, and was an active contributor to the H.264/MPEG-4 AVC standardization process (the '059 claims target arithmetic coding in the CABAC vein of that standard; the PTAB petition notes the JVT considered a Matsushita/Nokia-backed CAVLC proposal).
• Line of business: Global consumer-electronics and semiconductor manufacturer.
• Current status: Operating (Panasonic Holdings Corp.), but it sold this patent out of the operating company in 2014; it is no longer the owner.

Assignment timeline

Reel/frame numbers, execution vs. recordation dates, and correspondents are unverified — not retrievable from the sources available to me (Assignment Center search by patent number 7970059 at https://assignmentcenter.uspto.gov/ will show the cover sheets). Dates below are the event dates as published in the Google Patents legal-events log; conveyance types are quoted from that same log. Treat the two 2014 security-interest entries and the original-inventor assignment date with particular caution — see flags inline.

• ~2002–2003 (original, executed) / listed event date 2017-11-12 — Reel/frame unverified

  • Conveyance: Assignment of Assignor's Interest
  • Assignor: Abe, Kiyofumi; Hagai, Makoto; Kadono, Shinya; Kondo, Satoshi
  • Assignee: Matsushita Electric Industrial Co., Ltd.
  • Correspondent: unverified
  • Context: Original inventors-to-company assignment. Flag: the 2017-11-12 event date is anomalous — Matsushita renamed to Panasonic in 2008 — which suggests this entry is a later-recorded confirmatory/nunc-pro-tunc assignment of the divisional application (filed 2010) to clean up chain of title, not the original 2003 recordation. Unverified.

• 2014-01-24 — Reel/frame unverified

  • Conveyance: Lien (security agreement), recorded against grantor
  • Assignor: Optis Wireless Technology, LLC
  • Assignee: Highbridge Principal Strategies, LLC, as Collateral Agent
  • Correspondent: unverified
  • Context: Acquisition financing — the lender (Highbridge's principal-credit arm) took a lien on the just-acquired Panasonic portfolio days before the ownership assignment recorded.

• 2014-02-23 — Reel/frame unverified

  • Conveyance: Assignment of Assignor's Interest
  • Assignor: Panasonic Corporation
  • Assignee: Optis Wireless Technology, LLC
  • Correspondent: unverified
  • Context: Portfolio fire-sale / monetization — Panasonic transferred this patent (with a large H.264-era portfolio) to a licensing LLC, a transfer widely reported in 2014.

• 2014-03-12 — Reel/frame unverified

  • Conveyance: Security Interest
  • Assignor (grantor): Optis Wireless Technology, LLC
  • Assignee: Wilmington Trust, National Association
  • Correspondent: unverified
  • Context: Same financing structure — collateral agent/trustee security interest recorded ~2 weeks after the ownership transfer (standard asset-backed structure for the acquisition).

• 2016-07-15 — Reel/frame unverified

  • Conveyance: Release by Secured Party
  • Assignor (releasing party): HPS Investment Partners, LLC
  • Assignee (benefited party): Optis Wireless Technology, LLC
  • Correspondent: unverified
  • Context: Release of the Highbridge/HPS lien (HPS took over Highbridge's credit business in 2014), indicating the acquisition debt was refinanced or repaid.

• 2020-11-16 — Reel/frame unverified

  • Conveyance: Release by Secured Party
  • Assignor (releasing party): Wilmington Trust, National Association, as Collateral Agent
  • Assignee (benefited party): Optis Wireless Technology
  • Correspondent: unverified
  • Context: Release of the remaining security interest, clearing title ~5 weeks before the transfer to the litigation vehicle.

• 2020-12-22 — Reel/frame unverified

  • Conveyance: Assignment of Assignor's Interest
  • Assignor: Optis Wireless Technology, LLC
  • Assignee: VideoLabs, Inc.
  • Correspondent: unverified
  • Context: Transfer-to-asserter — the patent moved from the Optis licensing LLC to VideoLabs, the entity that has since filed the H.264 assertion campaign (Dell, Amazon, Meta, HP, Roku, TCL, Hisense, et al.).

Important caveat: The original inventors→Matsushita recordation and its 2017 event date could not be reconciled without the raw cover sheets. Confirm all dates (execution vs. recordation) and reel/frames directly at the USPTO Assignment Center before using this chart in any filing.

Timeline diagram

timeline
    title Ownership of US 7970059
    2003 : Filed as PCT by Matsushita
    2010 : Divisional application filed
    2011 : Patent issued
    2014 : Assigned to Optis Wireless
         : Highbridge lien recorded
         : Wilmington security interest
    2016 : HPS releases lien
    2020 : Wilmington releases interest
         : Assigned to VideoLabs Inc
    2021 : First infringement suit filed
    2023 : Patent expired by term

NPE / troll-pattern signals

1. Shell-entity transfer — present. The patent moved from an operating company (Panasonic, which shipped H.264 products) to Optis Wireless Technology, LLC (2014-02-23 event) and then to VideoLabs, Inc. (2020-12-22 event) — both licensing/assertion-only Delaware entities with no product business, classified as NPE by Unified Patents. Tells (no-product licensing LLCs; single-purpose assertion vehicles) are corroborated by their litigation conduct, not naming alone. Reel/frames unverified but the events are on the USPTO-derived record.

2. Known asserter in the chain — present. Unified Patents classifies VideoLabs Inc. as "NPE (Patent Assertion Entity)" in IPR2022-00701, and VideoLabs is a high-frequency plaintiff on the '059 patent (Dell 6:21-cv-00932; Meta 1:22-cv-00680; Amazon 6:22-cv-00079; HP 6:23-cv-00641; Roku 1:23-cv-01136; TCL 2:25-cv-00161; plus 2024–2026 E.D. Tex. cases). Optis Wireless Technology, LLC is likewise a known monetization/assertion entity (the Optis v. Apple litigation family). Both are on public NPE/asserter lists.

3. Repeat correspondent across the chain — unclear. I could not retrieve correspondent names from the USPTO Assignment Center or any indexed mirror within my search budget. This is the single most valuable missing datum for the "same lawyer runs the shell LLCs" tell, and it should be checked first at https://assignmentcenter.uspto.gov/ (search 7970059, open each cover sheet). No finding either way.

4. Cascading transfers — present. Four USPTO events cluster in ~7 weeks in early 2014 (2014-01-24 lien → 2014-02-23 ownership transfer → 2014-03-12 security interest), followed by two releases in 2016 and 2020, then the 2020-12-22 transfer to VideoLabs. This cadence — lien, assignment, security interest, later releases, then transfer to the litigation vehicle — is the signature of a lender-financed patent acquisition structured for enforcement.

5. Pre-litigation transfer — not present under the 6-month rule, but the timing is telling. The VideoLabs assignment is dated 2020-12-22; the first identified '059 infringement suit (VideoLabs v. Dell, W.D. Tex. 6:21-cv-00932) was filed 2021-09-10 — roughly 8–9 months later. Outside the 6-month window, but the transfer to the asserting entity immediately preceded the campaign, consistent with establishing a clean standing record.

6. Bankruptcy fire-sale — not present. Panasonic was not in bankruptcy; the 2014 transfer was a voluntary portfolio monetization (patent-backed financing), not a Chapter 7/11 asset sale.

7. Privateering — present (pattern, with a caveat). An operating company that practices and standardizes the technology (Panasonic/Matsushita, an H.264 contributor) sold the patent to a lender-backed NPE chain (Optis → VideoLabs) that now asserts it against H.264 implementers across the industry (Dell, Amazon, Meta, HP, Roku, TCL, Hisense). I found no public SEC filing quantifying a revenue-share or license-back to Panasonic, so I cannot confirm the exact economic arrangement — but the structure is the classic privateering template.

8. Defensive aggregator (anti-NPE) — not present. The chain terminates at VideoLabs, Inc., an assertor, not at RPX, AST, LOT, Unified Patents, or OIN. No neutralizing acquisition.

Verdict

NPE — high confidence

The record shows (i) a 2014-02-23 transfer from operating company Panasonic Corporation to the licensing LLC Optis Wireless Technology, LLC, wrapped in a lender-financed structure (2014-01-24 Highbridge lien; 2014-03-12 Wilmington Trust security interest; 2016 and 2020 releases), and (ii) a 2020-12-22 transfer from Optis to VideoLabs, Inc., an entity Unified Patents classifies as an NPE that has filed at least seven district-court campaigns naming this patent since 2021. Two strong signals — known asserter in the chain and transfer-to-assertion-vehicle — plus the cascading-financing pattern support high confidence; the only unverified element is the correspondent identity (reel/frame data), which I could not retrieve and which should be confirmed at https://assignmentcenter.uspto.gov/ (search patent 7970059) before any reliance on the chain details above.

The Unified Patents portal indicates US7970059 has "Patent Art (32)" and "Non-Patent Literature (6)" cited references. Let me pull the actual citation lists from multiple sources.

The freepatentsonline record surfaced the foreign citations and NPL. Now I need the complete U.S. patent citation list (backward citations) as recorded by Google Patents and the USPTO.

Prior Art Analysis — US Patent 7,970,059 B2

Important methodological caveats (read first)

1. USPTO access: In this session I could not directly query the USPTO Patent Center/PatFT "References Cited" image or the front-page citation data for 7,970,059. I relied on mirrors of USPTO data (Google Patents, freepatentsonline.com, Unified Patents, Justia) and on the IPR2022-00701 record. These are consistent with each other, but I could not confirm the complete U.S.-patent-portion of the examiner's citation list.
2. Claim text unavailable: The supplied full patent text is truncated before the formal "Claims" section, and my searches did not surface the verbatim claims of 7,970,059. Therefore, I cannot reliably map references to specific claim numbers. The § 102 mapping below is against the claimed subject matter as reconstructed from the specification's "Disclosure of Invention" (which mirrors the independent claims) — flagged as provisional. The patent's front-page citation count (per Unified Patents: "Patent Art (32)"; "Non-Patent Literature (6)") exceeds what I could enumerate, and the foreign/NPL lists below are the entries I could verify verbatim.
3. Distinguish citation directions: Lists found on Justia pages for later patents (e.g., 8,687,904; 9,258,558) show 7,970,059 as a forward citation (those patents cite '059). I have not counted those as prior art for '059.

Part 1 — References cited on the record of 7,970,059 (front-page / prosecution record)

Source: freepatentsonline.com/7970059.html (mirrors the USPTO front-page reference list); count corroborated by Unified Patents portal ("Patent Art (32), Non-Patent Literature (6)").

A. Foreign patent documents (verified on the record)

B. Non-patent literature (verified on the record)

C. Procedural documents on the record (not § 102 prior art)

• International Search Report, PCT/JP03/04804 (Jul. 15, 2003).
• Supplementary European Search Report, EP 03 720 906 (Aug. 26, 2004).
• European Office Action, EP 03 720 906 (Feb. 7, 2006).

Part 2 — Most relevant prior art asserted against the patent (IPR2022-00701, Dell v. VideoLabs)

The most probative prior-art references known for 7,970,059 come from the Dell IPR petition (filed 2022-03-17; not instituted on the merits, per Unified Patents/PTAB records). These are the references that Dell identified as the strongest anticipation/obviousness case against the '059 claims. Because the claims' text is unavailable to me, treat the § 102 mapping as an analytical framework rather than a verified claim-by-claim chart.

Part 3 — Consolidated § 102 analysis against the claimed invention

Because the claims text is outside my verified record, the following is mapped to the two reconstructed independent claims from the "Disclosure of Invention" (a coding method and a decoding method, each requiring: (i) scanning/ordering coefficient absolute values high-frequency → low-frequency; (ii) arithmetic coding/decoding of those absolute values while switching among plural probability tables; and (iii) switching in one direction only when a coded/decoded absolute value exceeds a predetermined threshold — i.e., no switch back down even if later values are smaller).

• Anticipation risk is concentrated on the "one-directional ratchet" + "threshold trigger" + "arithmetic coding of coefficient absolute values in high→low scan order" combination. The candidate art most on point for the arithmetic-coding half is the CABAC lineage (US 7,062,758; JVT-C060; H.264 draft JVT-D015d5), and for the table-switching-after-magnitude-trend half is H.263 Annex E (background) and Hartung et al. 1999 (multi-table VLC).
• H.263 Annex E discloses elements (i) and (ii) but the specification's own FIG. 2 shows bidirectional switching tied only to the immediately previous coefficient — it lacks the "in one direction / never back" limitation and the "once exceeded, stay" rule. On the reconstructed claim language, Annex E would not anticipate the independent claims; it is the closest § 102(b) starting point for an obviousness (§ 103) combination.
• Hartung et al. 1999 discloses multiple-table selection for DCT coefficient levels but uses VLC, not arithmetic coding, and no verified one-direction ratchet — partial only.
• EP0720379 / EP0873017 / EP1150434 / JP / KR documents are on the front-page record but I could not verify their disclosures this session; I will not assert anticipation against them beyond noting they are § 102(b)-age art (pre-AIA) whose subject matter (image entropy coding) makes them plausible cumulative references.
• US 7,062,758 / JVT-C060 / JVT-D015d5 — the strongest asserted anticipation/obviousness combination (per the IPR petition), with the central dispute almost certainly being (a) whether the claims are entitled to the April 2002 JP priority (affecting their § 102 status), and (b) whether the "one-directional switch after a threshold" limitation is disclosed or rendered obvious. The PTAB declined to institute IPR2022-00701, meaning the Board did not find a reasonable likelihood that Dell would prevail on those grounds — a useful (though not dispositive for a § 102 analysis) data point.

Part 4 — What I could not verify (and how to close the gap)

1. Exact claims (numbering and text) of 7,970,059 — needed for a defensible claim-by-claim § 102 chart. Verify via USPTO Patent Center (application 12/710,731) or the issued PDF at patents.google.com/patent/US7970059/en (Claims tab).
2. Complete U.S.-patent citation list on the '059 front page (the record shows 32 patent-art entries; I verified the foreign subset and NPL above but not the full U.S. subset).
3. Content of the EP0720379, EP0873017, EP1150434 and the JP/KR publications — titles only.

Recommended authoritative sources to finish the chart:

• USPTO Patent Center: https://patentcenter.uspto.gov (application 12/710,731 — "References Cited" and full claims)
• Google Patents citations tab: https://patents.google.com/patent/US7970059/en (scroll to "Patent Citations")
• IPR2022-00701 petition exhibits (USPTO PTAB API / Unified Patents PTAB portal) for Dell's full anticipation/obviousness grounds against each challenged claim.

I'll try to close two gaps that materially affect a § 103 analysis: the verbatim claims of the '059 patent, and the effective dates of the strongest asserted references.

Excellent — that search surfaced several pre-April-2002 references the earlier analysis didn't have. Let me try to pin down the claim text and the dates of the new CABAC/VLC references.

Obviousness Analysis — US 7,970,059 B2 (35 U.S.C. § 103)

Pre-AIA analysis. The '059 patent's priority chain runs to JP 2002-126029 (April 2002) via PCT/JP03/04804 (filed 2003-04-16) and application 10/481,277. Pre-AIA § 103 therefore applies, as construed through Graham v. John Deere and KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) (KSR applies to pre-AIA patents).

0. Corrections to the two earlier sections that drive this analysis

Two findings materially change the earlier work, and one of them reverses a strategic conclusion.

(a) The claim text is now recovered — and it is materially narrower than the "Disclosure of Invention" reconstruction. The earlier Patent summary correctly flagged that it could not verify claims. The Justia claim listing for patent 7970059 now supplies operative language, including:

"…(i) to switch between the plurality of probability tables in a predetermined one direction within each block such that each of the probability tables, which has been used for performing arithmetic coding on the first bit of the binary data corresponding to an already coded coefficient before switching to the new probability table, is not used within each block after switching to the new probability table, and (ii) not to switch between the plurality of probability tables in the direction opposite to the predetermined one direction regardless of the result of the comparison, and … within each block, if a predetermined one of the plurality of probability tables has been used, said first switching unit is configured not to switch between the plurality of probability tables regardless of the result of the comparison."

That snippet is the tail of claim 2; claim 3 is a decoding method reciting "receiving multiplexed data obtained by multiplexing (i) coded picture data … and (ii) audio data," demultiplexing, decoding the picture data, and decoding audio data; claim 4 recites "A semiconductor which executes the decoding method according to claim 3." Source: https://patents.justia.com/patent/7970059 (Patent History block confirms patent 7970059, app. 12/710,731, filed Feb 23 2010, date of patent Jun 28 2011, Panasonic). Caveat: I could not cross-verify this text against the USPTO Patent Center claim image; the claim set has features (first-bit-only switching, per-block table non-reuse, a terminal "predetermined one" table, a "semiconductor" claim) that are narrower and differently drafted than the specification's Disclosure of Invention. Verify numbering and text before filing anything.

Three limitations in that text are new relative to the earlier reconstruction and are the crux of any § 103 case:

• the switch governs only the first bit of the binary data for each absolute coefficient value;
• the "one direction" is defined as non-reuse of any previously used table within the block (a monotone ratchet, reset per block); and
• a terminal table — once the last table is reached, no further switching "regardless of the result of the comparison."

(b) The earlier Prior Art section's central strategic conclusion is too pessimistic. That section concluded that, per the Board's priority ruling, "post-April-2002 H.264/JVT art is a dead end" and that invalidity must rest on "H.263 Annex E, pre-2002 CABAC work, MPEG-4 and JVT documents dated before April 2002." The first half is right; the second half was left as a placeholder. That placeholder is now populated, and it is not empty. Live searching surfaced a family of pre-April-2002 JVT/VCEG documents that Dell never used — precisely because Dell built its petition on JVT-C060 (May 2002) and JVT-D015d5 (Aug 2002). These predate the priority date itself and therefore survive the Board's ruling unaffected by it.

Minor corroboration of bibliographic discrepancies already flagged: Justia independently states "Filed: Feb 23, 2010 / Date of Patent: Jun 28, 2011" (matching Google Patents), against Unified Patents' "2010-02-22 / 2011-06-27 / priority 2002-04-25 / expiration 2023-04-15." ±1-day variance persists across sources; nothing turns on it here.

1. The date matrix — this decides which combinations are even available

For pre-AIA purposes there are three zones. Under § 102(a), a printed publication need only predate the invention (presumptively the filing date, here at latest the 2002-04-26 JP priority date). Under § 102(b), the reference must predate the US filing date by more than one year; the '059's effective US filing date is 2003-04-16 via § 120 to PCT/JP03/04804, giving a critical date of ~2002-04-16. (The 2010 divisional filing does not move this.)

Key consequence: the obviousness case does not need to relitigate priority. A combination built entirely on Zone A art is a § 103 case the Board's priority holding does not touch.

Verification flag on the § 102(b) nuance: the weight of pre-AIA practice is that a § 119 foreign priority claim cannot be used to avoid a § 102(b) bar (measured from the US filing date); the leading § 119-limitation authority, In re Hilmer, 359 F.2d 859 (CCPA 1966), speaks to § 102(e). This does not matter for the Zone A references (all predate 2002-04-16 anyway), but have counsel confirm the point before it appears in a petition.

2. The claim elements to be met

Based on the recovered claim 2/3 language, the operative elements are:

1. (E1) Scan/order — coefficients of a two-dimensional array of frequency components, scanned "starting at a high frequency component toward a low frequency component."
2. (E2) Binarize — convert each coefficient absolute value into binary data.
3. (E3) Arithmetic code the first bit of that binary data using a plurality of probability tables.
4. (E4) Compare-and-switch — switch from the current table (for the first bit of coefficient n) to a new table (for the first bit of coefficient n+1) based on comparing |coefficient n| to a predetermined threshold.
5. (E5) One-direction / no-reuse — switching proceeds in a predetermined one direction within each block, so no table already used is reused within that block; never switch the opposite direction, regardless of the comparison result.
6. (E6) Terminal table — once the predetermined last table is used, no switching at all, regardless of the comparison result.
7. (E7) (claim 3) — receive multiplexed picture+audio data, demultiplex, decode picture data per E1–E6, decode the audio.
8. (E8) (claim 4) — a semiconductor that executes the claim 3 method.

3. Primary combination: H.263 Annex E + Hartung & Girod + JVT-B045

3.1 What each reference supplies

H.263 Annex E (1998) supplies E1–E4 nearly verbatim. The '059 specification itself admits this: frequency-transform a block, quantize, scan "starting at a high frequency component toward lower frequency components," form run/level pairs, binarize the level absolute values with a binary table, and arithmetic-code the binary data "by switching between a plurality of probability tables for use," where "[a] probability table is switched to another one based on a transition diagram shown in FIG. 2, depending on the absolute value of the previous coefficient value L" (tables 2/3/4 for previous |L| = 1, 2, ≥3). That is E1–E4 and the comparison logic of E4. What Annex E lacks is E5/E6 — it switches in both directions off the immediately previous coefficient, which the specification identifies as the very problem to be solved.

Hartung & Girod, "Improved Encoding of DCT Coefficients for Low-Bit-Rate Video Coding Using Multiple VLC Tables," ICIP 1999, pp. 51–55 (full text retrieved; https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1507415](/patent/1507415)/download-documents?artifactId=a5hG4FFHzR3hv8PM1p_tZ-RryDUkYx-2OqB4k6Spa8OpZisPanCIUzc) supplies the ratchet:

"We thus propose to use multiple VLC tables and to switch them according to the position in the block… We propose the idea to use a different VLC table T_n0 depending on the position of the last encoded/transmitted non-zero coefficient ȳ^Q_n0 when encoding the next non-zero coefficient… The index n0 of the previous non-zero coefficient, and thus of the VLC table, is known at the coder and at the decoder when the next coefficient is transmitted. Thus, no extra information needs to be signaled. Also, no additional computation is required… the tables T_n get smaller with increasing coefficient n0."

Because the table is indexed by the position of the last transmitted non-zero coefficient, that index advances monotonically through the block and never returns — Hartung's scheme is a one-direction, no-reuse table progression within a block, with a table set whose members are each used once. Hartung also supplies the factual predicate for why monotone selection is right: "the variances σ²(i,j) of DCT coefficients within blocks are known to decrease with increasing frequency," so "the statistics of the coefficients depend on the coefficient position in the block." His reported results: 8–18% bit-rate savings, up to 30% at fine quantization, and the "only drawback" being that "coder and decoder have to store more VLC tables."

JVT-B045 (Bjontegaard, RealNetworks/Telenor, generated 2002-01-23, presented at the JVT 2nd meeting, Geneva, Jan 29–Feb 1, 2002) supplies the bridge from VLC to arithmetic coding and the motivation to prefer cheap table management:

"TML specifies two different ways of entropy coding: — Use of one single VLC (UVLC) for all parameters… — CABAC (Context-based Adaptive Binary Arithmetic Coding). This method represents considerable coding gain (typically in the range (5–10)% reported in VCEG-O42). However, it is also significantly more complex… One low complexity and one high efficiency entropy coding method therefore seems to be a good solution."

JVT-B045 also discloses adaptive VLC-table selection driven by previously coded magnitude information, using "How many of the 4×4 blocks (a,b,c,d) have coefficients (M)" and "How many coefficients the last coded 4×4 block had (N)," with the rule "VLC2: M = 2 or M > 2 and N < 3; VLC3: M > 2 and (N = 3 or N = 4); VLC4: M > 2 and N > 4." That is threshold-vs-previously-coded-magnitude table selection with a saturating terminal table (VLC4 is the last; beyond N > 4 nothing changes) — structurally the E5/E6 ratchet, expressed for a different table type. (Retrieved text and the accompanying expert-declaration excerpts: same PTACTS petition-exhibit URL; docket corroboration of its public availability at https://www.courtlistener.com/docket/[4612948/321](https://assignmentcenter.uspto.gov/search/patent/reelFrameDetail?reelFrame=4612948-0321)/optis-wireless-technology-llc-v-huawei-device-shenzhen-co-ltd/ — JVT-B045 appears there as DX 199.)

3.2 Element-by-element

The only element the primary combination does not supply is E3's "first bit of the binary data" as the sole object of the switching decision — and that is addressed in § 4.

3.3 Motivation to combine (KSR factors)

1. Same field, same problem, same solution class. All three references address entropy coding of quantized transform coefficients in block-based video codecs. H.263 Annex E and JVT-B045 are literally the same standard family (H.263 → H.26L/TML); Hartung implemented his method "into a H.263 video coder."
2. The problem is named and quantified in the art. Hartung documents that coefficient statistics are non-uniform and position-dependent, with variances decreasing with frequency; the '059 specification states the identical premise in reverse ("the absolute values of coefficients … tend to be larger in the direction from higher frequency components toward lower frequency components") and identifies the deficiency of bidirectional switching ("the switching of a probability table cannot support an increase trend of coefficients").
3. A finite, identified, predictable set of solutions. Given three table-management policies — switch off the immediately previous coefficient (H.263), latch monotonically (Hartung), or saturate at a terminal table (JVT-B045) — selection among them is "a finite number of identified, predictable solutions" (KSR at 421). Nothing in H.263 Annex E criticizes or disparages latching, so there is no teaching-away.
4. Predictable result with an articulable mechanism. Arithmetic coding gains when the symbol probabilities in a table are more skewed; a table whose adaptation counts are not periodically diluted by a return to a shallower model will adapt more sharply. The patent's own Disclosure of Invention asserts exactly this as the consequence — which is a concession that the benefit follows mechanically, weighing against any "unexpected results" argument.
5. Design incentive the art states expressly. Hartung: the "only drawback" of multiple tables is storing more of them, with no signaling cost; JVT-B045: the low-complexity/high-efficiency trade-off is an explicit standardization goal. Reducing table count and switch count (which latching does) is an identified cost-reduction objective.
6. Standardization pressure. The TML/H.26L participants were competing to improve entropy-coding efficiency in the same window; the '059 family's own sibling filings (below) show the same assignee was doing precisely this.

3.4 A non-art data point that corroborates motivation (flag: not § 102 art)

A German-language family/sibling document, DE60318918T2, retrieved during this search, describes switching between VLC table 1 and VLC table 2 (not probability tables) where "wenn ein absoluter Wert von L einen Schwellwert übersteigt, wird die VLC-Tabelle 2 für die folgenden RL-Werte verwendet" and then, when |L| falls back below the threshold, "wird jedoch nicht zu der VLC-Tabelle 1 gewechselt" — "[d]ies bedeutet, dass eine Wechselrichtung zwischen den Tabellen eindirektional ist." It gives the identical rationale as the '059: |L| tends to increase as coefficients are one-dimensioned from high to low frequency; frequent switching is avoided; and the working-memory/ROM-expansion overhead of switching tables is reduced. The prioritised drafting it reflects — a Panasonic/Matsushita application identifying Kiyoshi/F. Abe, Kadono, and Hagai and filed 2002-04-15 (JP 2002-112665, "Methods of coding and decoding pictures"), whose claim 1 recites a "table switching step of switching tables to be referred to when the variable length coding has been performed" — is dated eleven days before the '059's JP priority date, with three of four inventors in common.

Treat this as corroborative evidence of what was obvious, not as prior art. It almost certainly fails "by another" under § 102(a)/(e), and its JP priority date cannot be carried into a § 102(e) date (Hilmer). Raise it only as a secondary-considerations rebuttal (simultaneous/parallel development by the same skilled artisans).

4. The "first bit" limitation (E3) — highest-upside combination, content unverified

E3 (switching applies to "the first bit of the binary data corresponding to each absolute value") is the limitation most likely to be used to distinguish the art. It should not be. In the '059's own binarization (FIG. 1: |L| = 2 → "01", = 3 → "001", = 1 → "1"), the first bit is the unary terminator/significance flag — a "greater-than-1?" indicator — and the specification's own variation affirmatively builds around that: with a threshold of "1," tables 1 and 4 only, the specification argues that "binary data is represented only as '1' when the absolute value of a coefficient value L is '1'. Therefore, when '1' is set as a threshold value, adaptations are made to the probability table 1 … so that a symbol (binary data) '1' occurs with a high probability."

Known CABAC practice independently located the model-selection signal at that same first bin: level coding binarizes |level| as a first bin distinguishing "1" from "greater than 1," with the context for that first bin derived from how many previously coded coefficients in the block exceeded 1. Because that count only increases within a block, the first-bin context index advances monotonically, is never reused, and saturates at a maximum context beyond which it does not change — E3 + E4 + E5 + E6 in a single mechanism, with a per-block reset.

Verification is required before relying on this route. The document that is publicly indexed as containing this content is JVT-C060 ("Improved CABAC," Schwarz et al., May 2002) — Zone B, and therefore unavailable under the Board's priority ruling. The pre-April-2002 counterparts exist and were publicly available — VCEG-O18 "Improved CABAC," VCEG-O24 "Improvements on CABAC" (~Nov 2001), and JVT-B101 "New Results on Improved CABAC" (Jan 2002) are all listed as standard contributions on Marpe's own HHI page (https://iphome.hhi.de/marpe/cabac.html) and appear as trial exhibits DX 205, DX 206, and DX 202 in Optis Wireless Technology, LLC v. Huawei Device (Shenzhen) Co., Ltd., E.D. Tex. 2:17-cv-00123, Docket 321 (https://www.courtlistener.com/docket/[4612948/321](https://assignmentcenter.uspto.gov/search/patent/reelFrameDetail?reelFrame=4612948-0321)/optis-wireless-technology-llc-v-huawei-device-shenzhen-co-ltd/). I did not retrieve their text and will not assert their contents. The statement "CABAC existed and delivered 5–10% gains before the priority date" is, however, independently grounded: JVT-B045 (Jan 2002) says so and cites VCEG-O42.

Proposed combination 3 (highest upside): VCEG-O18 / VCEG-O24 / JVT-B101 (CABAC: binarize |level|, switch context models per bin, first bin flagged against a threshold of 1, context advanced by the count of prior levels > 1 and saturated at a terminal context) + H.263 Annex E (plural probability tables switched off the previous coefficient's absolute value, high→low scan) + Hartung 1999 (monotone, no-reuse, non-signaled table progression). If those VCEG documents disclose the monotone/saturating first-bin context, the case escalates from a strong § 103 combination to a serious § 102 anticipation argument.

5. Claims 3 and 4 (E7, E8)

These add nothing of independent inventive weight:

• E7 (receive multiplexed picture+audio, demultiplex, decode video per E1–E6, decode audio) is the conventional container/systems layer. The '059's own face cites ISO/IEC 14496-1 (MPEG-4 Systems) — Dell's Ex. 1004 — and MPEG-4/H.324-style multiplexing long predates the priority date. The '059 specification's FIG. 16 describes precisely such a multiplexing/demultiplexing unit (ex 308) in a cellular phone. The motivation is the ordinary one of delivering video-with-audio in a system, plus the same efficiency gain in the video path.
• E8 ("a semiconductor which executes the decoding method according to claim 3") is an apparatus/hardware claim to a routine implementation choice. The specification itself describes implementation "as an LSI ex 117"; implementing a defined decoding algorithm in an LSI was standard practice. A claim whose only added element is "do it in silicon" does not carry independent non-obviousness weight where the underlying method is obvious (In re O'Farrell — expected/predictable variation).

6. The patentee's likely defenses, and how they fare

7. Bottom line

Strength: moderate-to-strong, and materially stronger than the earlier Prior Art section implied — but claim-mapping on the recovered narrow claim text is the gating work.

• Best combination: H.263 Annex E + Hartung & Girod (ICIP 1999) + JVT-B045 (Jan 2002), with VCEG-O18/O24/JVT-B101 added for the first-bin/level-context element. Every member is pre-priority-date Zone A art, so the IPR2022-00701 institution denial — which turned entirely on May/August 2002 references — does not dispose of this combination, and no § 315(e) estoppel exists because no IPR ever resulted in a final written decision.
• Gaps to close before filing or arguing:
  1. Pull VCEG-O18, VCEG-O24, JVT-B101, JVT-B002d2, and JVT-B015 from the ITU-T AV-arch/JVT archive (the Optis v. Huawei exhibit list shows the first three were obtainable and were used as trial exhibits) and determine whether the coefficient level coding context is derived from the count of prior levels > 1 and saturates.
  2. Verify the verbatim claims at USPTO Patent Center (app. 12/710,731) — including whether the "semiconductor" claim really is claim 4, and whether an apparatus independent claim exists.
  3. Confirm the H.263 Annex E text for the plural probability tables, the previous-coefficient-based switching, and the high-frequency-first scan, rather than relying on the '059's characterization of it.
  4. Confirm the § 102(b) critical-date nuance for Zone A/B line-drawing.
• Posture: the patent expired 2023-04-16 (status: Expired – Lifetime), so invalidity now matters only as a defense to pre-expiry damages, and the proof standard in district court is clear and convincing evidence (Microsoft v. i4i). A new IPR is possible in principle against pre-April-2002 art but is uneconomic for a 2023-expired patent outside a pending case; check § 315(b) timing if you are a served defendant.

Sources: Google Patents full text of US7970059 (authoritative, as supplied); Justia claim listing for patent 7970059 (https://patents.justia.com/patent/7970059); Hartung & Girod, ICIP 1999 full text and JVT-B045 full text via https://ptacts.uspto.gov/ptacts/public-informations/petitions/1507415/download-documents?artifactId=a5hG4FFHzR3hv8PM1p_tZ-RryDUkYx-2OqB4k6Spa8OpZisPanCIUzc ; Optis Wireless Technology v. Huawei Device (Shenzhen), E.D. Tex. 2:17-cv-00123, Dkt. 321 exhibit list, https://www.courtlistener.com/docket/[4612948/321](https://assignmentcenter.uspto.gov/search/patent/reelFrameDetail?reelFrame=4612948-0321)/optis-wireless-technology-llc-v-huawei-device-shenzhen-co-ltd/ ; Marpe HHI publication list, https://iphome.hhi.de/marpe/cabac.html ; IEEE Xplore record for Hartung & Girod (https://ieeexplore.ieee.org/abstract/document/[822853](/patent/822853)); DE60318918T2 machine text via Google Patents; Unified Patents portal for 7970059 and IPR2022-00701 (https://portal.unifiedpatents.com/[ptab](/ptab)/case/IPR2022-00701); freepatentsonline.com/7970059.html.

Caveats: (i) I could not verify the '059 claim text against Patent Center, and the recovered text differs materially from the earlier reconstruction — re-verify before reliance; (ii) I did not retrieve the text of VCEG-O18, VCEG-O24, JVT-B101, JVT-B015, or JVT-B002d2, and assert nothing about their specific disclosures; (iii) the DE60318918T2 / JP 2002-112665 material is treated as non-art corroboration only; (iv) the pre-AIA § 102(b)/§ 119 foreign-priority point should be confirmed by counsel.