Perceptual Quality Assessment Patents: Leaders & Trends 2026
- Filing is concentrated at the top: the top 5 assignees hold 27.6% of all 2,731 records in scope, and the top 10 hold 40.1% — a long tail of single- and few-filing entrants fills the rest.
- Growth has cooled from its 2020 peak: filings fell from 211 in 2021 to 112 in 2024, a -47% swing over that span, though 2025-2026 figures are still filling in as publication catches up with filing.
- Claims cluster in transmission, not media processing: H04L digital transmission covers 55.9% of records and H04W wireless networks 32.8%, while G06T image processing sits at just 5.0% — the network layer is crowded, the perceptual-modeling layer is comparatively open.
Filing growth compares 2021 (211 records) with 2024 (112) — a three-year span. 2024 is the most recent year we treat as complete: publication lags filing by roughly 18 months, so 2025 onwards are still filling in and any growth rate that ends there would understate the field. Top-5 share is the combined record count of the five largest assignees divided by all 2,731 records in scope (CR5), not by the ranked leaders only.
What this landscape covers
Perceptual quality assessment patents span the methods used to estimate how a human viewer or listener experiences degraded media — mean opinion score prediction, no-reference quality metrics, banding and blocking artifact detection, temporal pooling of frame-level scores, just-noticeable-difference thresholds, and per-title encoding decisions driven by those estimates. The dataset covers 2,731 published records filed or published between 2015 and the 2026-07-31 cut-off, drawn from a search string that pairs quality-of-experience terminology with the specific artifact and scoring vocabulary examiners use to classify these filings.
Most of this activity sits inside telecom and network infrastructure filings rather than pure video-codec research: the IPC composition below shows transmission and wireless classes carrying far more records than image-processing classes, which tells a filer where claim space is already dense and where it is not.
Filing trends and technology composition
The record set runs from 2015 through the 2026-07-31 cut-off; the two views below show when filings happened and which IPC subclasses they were classified under.
A 2020 peak followed by a real but partial decline
Filings rose from 162 in 2017 to a peak of 248 in 2020, then declined; measured on complete years, 2021's 211 filings fell to 112 by 2024, a -47% change. 2025 and 2026 counts are shown for completeness but understate true filing activity, since publication typically lags filing by around 18 months.
Transmission classes dominate over media-specific ones
H04L (digital information transmission) appears in 55.9% of the 2,731 records and H04W (wireless networks) in 32.8%, versus 25.7% for H04N (pictorial communication) and just 5.0-7.6% for the AI, image-processing and general-computing classes — G06N, G06T and G06F. Because a single record can carry several IPC codes, these shares add to well over 100% and should be read against the 2,731-record total, not against each other.
Shares are the percentage of the 2,731 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Perceptual Quality Assessment with Eureka
This page is one run against one query. Ask Eureka your own question about perceptual quality assessment and every answer comes back with the patent numbers behind it.
Try EurekaMost-cited records and a representative filing
US9813523B2 — Apparatus, method and system of quality of experience indication
Filed by Intel, this patent covers a radio configured to communicate encoded multimedia data over a wireless channel together with a QoE estimator that derives a quality-of-experience indicator from a mean opinion score for the encoded data plus a channel quality parameter, then feeds that indicator to a session manager.Granted 2017-11-07.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20120311126A1 | Systems and methods for measuring quality of experience for media streaming | 276 |
| 2 | US20200112907A1 | Quality of service information notification to user equipment, users, and application server | 275 |
| 3 | US20130182700A1 | Systems and methods for network monitoring and testing using a generic data mediation platform | 237 |
| 4 | US20140149557A1 | Network-Capacity Optimized Adaptive HTTP Streaming | 228 |
| 5 | US20090034426A1 | Monitoring quality of experience on a per subscriber, per session basis | 228 |
| 6 | US20190222489A1 | NETWORK DATA ANALYTICS FUNCTION (NWDAF) INFLUENCING FIFTH GENERATION (5G) QUALITY OF SERVICE (QoS) CONFIGURAT… | 216 |
| 7 | US20100061446A1 | Video signal encoding | 205 |
| 8 | US20140269269A1 | System and Methods for Estimation and Improvement of User, Service and Network QOE Metrics | 194 |
| 9 | US20150113133A1 | System and method for observing and controlling a programmable network using closed loop control | 185 |
| 10 | US20140071895A1 | Network Selection Based On Customizing Crowdsourced Connection Quality Data | 184 |
Citation counts inside this corpus favour older filings that have had more time to accumulate citations; treat them as a signal of influence on the field, not of current commercial weight.
Each row carries its publication number; clicking a row searches Eureka by that number.
Put your own technology through the same analysis
Eureka on the web
When you want the answer in the next five minutes.
The agent works the prompt against patents and technical literature, citing every source.
Run your analysis now →MCP server & REST API
When it has to run inside your own pipeline.
Patent search, landscape analysis and assignee resolution as MCP tools. Drop them into any agent framework, or call REST directly.
Browse MCP servers →What the numbers mean for a filing decision
Three patterns in the data matter more than the raw counts: where filings concentrate, where they are pulling back, and where the classification data shows open ground.
The leader board is short, the tail is long
A handful of large telecom and infrastructure filers hold well under half of all records between them, meaning the remaining roughly 60% is spread across a long tail of smaller and single-filing entrants. That tail is where freedom-to-operate work usually turns up unexpected blockers, because it is under-indexed in most competitor-watch lists.
Growth has genuinely slowed, not just paused
Filing counts fell across three consecutive complete years rather than dipping once and recovering, which is a stronger signal than a single down year. Recent-year momentum figures by individual assignee show declines of -50% to -100% year-on-year at several of the largest filers, consistent with the aggregate trend.
The perceptual-modeling layer is comparatively open
Transmission and wireless classes carry the bulk of filings, but the classes that cover actual perceptual modeling and AI-based quality estimation sit in single digits. That gap does not mean the technology is unclaimed, but it does mean the claim density a filer would compete against there is far lower than in the network layer.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to perceptual quality assessment, with the prior art for and against each one.
Who is filing, and where the gaps sit
The ranked leaders are dominated by network equipment and telecom operators rather than pure media-technology firms, which shapes what their claims tend to protect: session and channel-quality indication more often than perceptual modeling itself.
A clear single leader ahead of the field
The top-ranked assignee holds 262 records, well ahead of fifth place at 86 and tenth place at 59 — a steep drop-off rather than a gradual one. That gap is worth noting when scoping a competitor watch: the leader's portfolio breadth is not matched by anyone else in the ranking.
Co-filing is limited and concentrated in one relationship
Only ten co-assignee pairs appear across the dataset, and the strongest of them accounts for a meaningfully larger share of joint filings than the others. Most assignees in this field file alone rather than through joint ventures or co-development arrangements.
Even the leaders are pulling back
Momentum figures for several of the largest assignees show sharp year-on-year declines in the latest year, with more than one filer at zero. This is consistent with the aggregate -47% multi-year trend rather than an artefact of one company's strategy shift.
| Assignee | Recent year | YoY |
|---|---|---|
| Huawei Technologies Co., Ltd. | 2 | -67% |
| Samsung Electronics Co., Ltd. | 1 | -83% |
| Verizon Patent and Licensing Inc. | 1 | -50% |
| Telefonaktiebolaget LM Ericsson (publ) | 0 | -100% |
| Cisco Technology, Inc. | 0 | — |
| Lenovo (Singapore) Pte. Ltd. | 0 | -100% |
| Nippon Telegraph and Telephone Corporation | 0 | — |
| Digimarc Corporation | 0 | -100% |
Where to take this next
The landscape data points to specific follow-up questions rather than a single conclusion.
Check freedom-to-operate against the long tail
With the top 10 holding only 40.1% of records, most of the field's claim density sits outside the largest assignees. A freedom-to-operate review that only checks the ranked leaders will miss most of the corpus.
Run a freedom-to-operate scan in Eureka →Map claim language in the under-claimed classes
G06T, G06N and G06F carry far fewer records than the transmission classes, but that gap needs verifying at the claim level before treating it as open ground.
Explore claim charts in Eureka →Watch whether the 2021-2024 decline continues
Three consecutive years of falling filings is a real trend, but 2025-2026 data is still incomplete. Revisit this once publication catches up.
Track filing trends in Eureka →Common questions about this landscape
The dataset's assignee ranking shows a single leader with 262 records, well ahead of the rest of the ranked field, where fifth place holds 86 and tenth place 59. This is a steep drop rather than a gradual one, meaning the leader's portfolio breadth is not closely matched by any other single filer. Most of the remaining activity is spread across a long tail of the 100 ranked companies, many with only a handful of records each, so a competitor watch focused only on the largest names will miss most of the field.
Filing peaked in 2020 at 248 records and has declined since on complete-year data: 211 filings in 2021 fell to 112 by 2024, a -47% change over that span. That is a genuine multi-year decline rather than a single down year. Figures for 2025 and 2026 are included in the raw trend but should not be read as confirming further decline, because publication typically lags filing by roughly 18 months and those years are still filling in.
The two largest classes are H04L (digital information transmission), covering 55.9% of the 2,731 records in scope, and H04W (wireless communication networks) at 32.8%. H04N (pictorial communication, i.e. video and TV) covers 25.7%, while classes tied more directly to perceptual modeling and AI-based estimation — G06T, G06N and G06F — each sit between 5.0% and 7.6%. Because records often carry multiple IPC codes, these percentages add to more than 100% and should each be read against the full 2,731-record total.
US9813523B2, assigned to Intel and granted 2017-11-07, covers an apparatus with a radio for communicating encoded multimedia data over a wireless channel, paired with a quality-of-experience estimator. That estimator derives a QoE indicator from a mean opinion score for the encoded data combined with a channel quality parameter, then passes the indicator to a session manager. Anyone building a wireless streaming system that derives session-level QoE from MOS plus channel conditions in this specific combination should review this claim scope closely before finalising an architecture.
The IPC composition data shows transmission and wireless classes carrying the bulk of filings, while classes covering perceptual modeling itself — image data processing (G06T), AI-based computing models (G06N) and general digital data processing (G06F) — each cover only 5.0% to 7.6% of the 2,731 records. Sub-areas like no-reference banding artifact detection, temporal pooling of frame-level scores, and just-noticeable-difference thresholding for encoding decisions sit adjacent to this lighter-filed territory. This does not guarantee the space is unclaimed, but it does mean claim density there is markedly lower than in the network-layer classes.
Research Perceptual Quality Assessment in depth with Eureka
Go past this page: query the whole perceptual quality assessment corpus yourself, in your own scope.
Every answer comes back with patent numbers you can open.
Disclaimer. This page is generated from Patsnap Eureka data drawn from a limited snapshot of global patent and scientific-literature records, and is provided for general information and reference only.
Patent data carries inherent limitations: recent filings (typically the most recent 18–24 months) are under-counted due to standard publication lag; counts may be reported at either a patent-family or a patent-record basis and are not always directly comparable; classification, applicant-name, and citation data may contain errors, duplicates, or omissions; and the underlying search query defines and constrains the scope shown. As a result, the analysis may be incomplete or inaccurate and may not reflect the full technology landscape.
Nothing on this page constitutes an exhaustive prior-art, novelty, freedom-to-operate, or validity search, nor does it constitute legal, financial, investment, or professional advice, and it should not be relied upon as such. Any patent, commercial, or strategic decision should be verified independently and reviewed with qualified patent, legal, and domain professionals. Patsnap makes no warranties, express or implied, as to the accuracy, completeness, or fitness for any particular purpose of the information presented.
Machine translation. Assignee and organisation names originally recorded in Chinese, Japanese or Korean have been rendered into English by an AI translation step so that the tables stay readable. These renderings are best-effort and may not match a company’s registered English name; the original name is what the underlying patent record carries, and it is what any Eureka query launched from this page uses.