Could a single patent shake up the way digital audio platforms operate? That’s exactly what is happening with patent US11899713B2, which has become the focus of multiple lawsuits filed by Muvox LLC, a non-practicing entity (NPE). Major companies like Spotify, Meta, and Netaktion are now facing litigation over this patent, and the legal battle is far from over.
At its core, US11899713B2 relates to adaptive audio streaming and personalized content delivery. Imagine you are streaming music or a podcast. This technology ensures that the audio quality dynamically adjusts based on your internet speed, device capabilities, and even your listening habits. It is designed to create a seamless, uninterrupted experience, reducing buffering and optimizing playback.
This patent is being aggressively asserted because adaptive streaming is a fundamental part of how modern digital audio services function. Whether it is Spotify personalizing your music recommendations or Meta optimizing audio in its platforms, the underlying technology has become essential in the industry. That’s why Muvox LLC is targeting multiple companies, claiming that their platforms infringe on its patent rights.
With five active cases, two inactive ones, and even a re-examination underway, this litigation could have major implications for streaming services. But is this patent truly unique? Or are there prior patents that could challenge its validity?
That’s where Global Patent Search (GPS) comes in. By analyzing feature mappings and historical patents, GPS helps uncover related patents that potentially invalidate claims and offer a stronger legal defense.
Understanding Patent US11899713B2
US11899713B2, owned by Muvox LLC, covers a system and method for music streaming, playlist creation, and track categorization. The patent describes a way to analyze music tracks based on Rhythm, Texture, and Pitch (RTP) scores, which are derived from the track’s low-level acoustic data. These RTP scores help categorize music into different moods and genres, making it easier for music publishers to generate playlists and stream tracks.
A key aspect of US11899713B2 is that once a track has been RTP scored, the data is stored in a universal database, allowing multiple Muvox LLC-licensed publishers to use the same categorization without redundant processing. The system operates through an API server, enabling end-user applications to import categorized tracks, create playlists, and initiate streaming, all within a publisher-controlled environment.
Source: Google Patents
Its four key features are:
#1. Automated music categorization: The system assigns RTP scores to tracks, classifying them based on rhythm, texture, and pitch to generate intelligent playlists.
#2. Universal database of RTP scores: Once analyzed, a track’s RTP score is stored in a shared database, preventing multiple publishers from re-processing the same track.
#3. API-based playlist and streaming control: Music publishers can use an API server to import, organize, and stream tracks for their users.
#4. Publisher-specific access: End-user applications can only access tracks from their sponsoring publisher, ensuring licensing compliance and monetization for Muvox LLC.
Since playlist creation and streaming architectures are critical components of modern music platforms, Muvox LLC is leveraging US11899713B2 to seek licensing fees or settlements from these companies. However, a key question remains: Are there earlier patents that describe similar playlists and streaming systems? Let’s explore possible related patents that could challenge the patent’s uniqueness.
Related Patent References for US11899713B2
#1. US20220245193A1
US20220245193A1, filed on April 14, 2022, by Aperture Investments LLC, describes a system and method for music streaming, playlist creation, and track categorization. The invention focuses on analyzing songs using computer-derived Rhythm, Texture, and Pitch (RTP) scores, which are used to categorize tracks and create mood-based playlists.
Source: GPS
Key Features of this Related Patent:
- Selecting a song based on a computer-derived comparison – The reference describes a system that categorizes tracks using RTP scores, enabling automated playlist creation based on song similarities.
- Use of a human-trained machine for song analysis – The reference states that neural networks trained on human-labeled sound clips are used to analyze and classify songs.
- Mapping frequency characteristics to moods – RTP scores are mapped to moods such as happy, sad, and excited, which directly correlates to playlist creation techniques in US11899713B2.
- Selection of songs based on mood similarity – The system enables dynamic playlist generation by comparing the moods of different songs, aligning with US11899713B2’s feature of selecting tracks based on mood similarity.
Which features of US11899713B2 are disclosed by US20220245193A1?
| Key Feature of Claim | Disclosure Status |
| Selecting a song based on computer-derived comparison between song representation and known similarities in other song representations | Fully Disclosed |
| Known similarities in song representations are based on a human-trained machine | Fully Disclosed |
| Song representation is based on isolating and identifying frequency characteristics | Fully Disclosed |
| Representations of other songs are based on human listening to isolate and identify frequency characteristics | Fully Disclosed |
| Frequency characteristics correspond to moods of the songs | Fully Disclosed |
| Selection is based on similarity between moods of the song and moods of other songs | Fully Disclosed |
Key Excerpt from US20220245193A1:
“Embodiments of the present disclosure are primarily directed to music categorization, playlist creation, and music streaming to an end-user application. In particular, embodiments involve a music categorization system that objectively categorizes music based on rhythm, texture, and pitch (RTP) values or scores, from which the mood or some other category of the music may be determined and used to create playlists.”
#2. US20200228596A1
US20200228596A1, filed on April 1, 2020, by Aperture Investments LLC, describes a streaming music categorization system using Rhythm, Texture, and Pitch (RTP) scores. The invention focuses on analyzing low-level and high-level frequency data to classify songs into moods and categories.
Key Features of this Related Patent:
- Computer-based song selection using frequency data – The system categorizes songs using RTP scores, allowing for automated playlist generation.
- Extraction of frequency characteristics for song representation – The reference describes using spectrograms and other analysis methods to isolate key frequency elements in music.
- Mapping frequency characteristics to moods – RTP scores are linked to moods such as happy, sad, and calm, aligning with US11899713B2’s playlist generation approach.
- Selection based on mood similarity – The system compares RTP scores to categorize songs with similar emotional tones, matching a core functionality of US11899713B2.
Which features of US11899713B2 are disclosed by US20200228596A1?
| Key Feature of Claim | Disclosure Status |
| Selecting a song based on computer-derived comparison between song representation and known similarities in other song representations | Fully Disclosed |
| Known similarities in song representations are based on a human-trained machine | Partially Disclosed |
| Song representation is based on isolating and identifying frequency characteristics | Fully Disclosed |
| Representations of other songs are based on human listening to isolate and identify frequency characteristics | Partially Disclosed |
| Frequency characteristics correspond to moods of the songs | Fully Disclosed |
| Selection is based on similarity between moods of the song and moods of other songs | Fully Disclosed |
Key Excerpt from US20200228596A1:
“Values for RTP may be determined holistically, based on low-level data extracted from the music, or high-level data constructed or derived from the low-level data. An example of a holistic method for determining RTP is analyzing frequency-related data from tracks using spectrograms, isolating and categorizing tracks based on similarities in rhythm, texture, and pitch.”
#3. US20150268800A1
US20150268800A1, filed on October 14, 2014, by Timothy Chester Okonski and Masahiko Nishimura, describes a method and system for dynamic playlist generation. The invention focuses on matching user conditions with pre-classified song metadata, such as mood categories and pacing levels, to generate personalized playlists.
Key Features of this Related Patent:
- Metadata-based song classification for playlist generation – The system categorizes songs using metadata such as mood and pacing levels to create personalized playlists.
- User preference matching for automated playlist creation – Playlists are dynamically generated based on user-selected mood categories.
- Mood-based categorization of songs – Songs are tagged with mood categories, and users can filter and select music accordingly.
- Selection based on mood similarity – The system creates personalized playlists by comparing a user’s selected mood with pre-assigned mood classifications of songs.
This is how feature mapping from the tool looks like:
Source: GPS
Which features of US11899713B2 are disclosed by US20150268800A1?
| Key Feature of Claim | Disclosure Status |
| Selecting a song based on computer-derived comparison between song representation and known similarities in other song representations | Partially Disclosed |
| Frequency characteristics correspond to the moods of the songs | Partially Disclosed |
| Selection is based on the similarity between the moods of the song and the moods of other songs | Fully Disclosed |
Key Excerpt from US20150268800A1:
“In a preferred embodiment, the classifications include categories of moods. The personalized playlist can be generated based on matching the mood selections of the user with the mood categories assigned to the media tracks.”
#4. EP2843860A1
EP2843860A1, filed on August 26, 2013, by Panasonic Automotive Systems Company of America, describes a method and system for preparing a playlist for an internet content provider. The invention focuses on automated playlist generation based on similarity coefficients and metadata descriptors, which help classify and group songs.
Key Features of this Related Patent:
- Metadata-based playlist generation – Songs are categorized using track-to-track similarity coefficients and artist-to-artist metadata.
- Software-generated song similarity calculations – The system uses computer-generated similarity data to compare tracks.
- Mood-based song classification – Mood annotations and classifications are used in playlist formation.
- Selection based on similarity descriptors – The reference describes playlist definitions that use similarity data descriptors to group songs.
Which features of US11899713B2 are disclosed by EP2843860A1?
| Key Feature of Claim | Disclosure Status |
| Selecting a song based on computer-derived comparison between song representation and known similarities in other song representations | Partially Disclosed |
| Song representation is based on isolating and identifying frequency characteristics | Partially Disclosed |
| Frequency characteristics correspond to moods of the songs | Partially Disclosed |
| Selection is based on similarity between moods of the song and moods of other songs | Partially Disclosed |
Key Excerpt from EP2843860A1:
“The metadata encyclopedia may include computer-software-generated track-to-track similarity coefficients; computer-software-generated artist-to-artist similarity coefficients; explicit user preference information; listener’s past selections, and mood-based annotations.”
#5. WO2010027509A1
WO2010027509A1, filed on September 8, 2009, by Sourcetone LLC, describes a music classification system and method. The invention focuses on analyzing digital music to classify songs based on emotional responses and frequency characteristics.
Key Features of this Related Patent:
- Machine learning-based emotional classification of music – Songs are analyzed using audio signal features and human-trained machine learning models.
- Extraction of frequency characteristics for classification – The system uses spectral processing and frequency analysis to isolate key audio elements.
- Mood-based song categorization – Emotional responses to music are mapped to mood-based classifications such as happy, sad, and anxious.
- Selection based on mood similarity – Songs are grouped and recommended based on emotional similarity metrics.
Which features of US11899713B2 are disclosed by WO2010027509A1?
| Key Feature of Claim | Disclosure Status |
| Selecting a song based on computer-derived comparison between song representation and known similarities in other song representations | Partially Disclosed |
| Known similarities in song representations are based on a human-trained machine | Fully Disclosed |
| Song representation is based on isolating and identifying frequency characteristics | Fully Disclosed |
| Representations of other songs are based on human listening to isolate and identify frequency characteristics | Partially Disclosed |
| Frequency characteristics correspond to moods of the songs | Fully Disclosed |
| Selection is based on similarity between moods of the song and moods of other songs | Fully Disclosed |
Key Excerpt from WO2010027509A1:
“An audio analysis process derives descriptive features from an audio input signal representing digitally recorded music. Emotional classifications are developed using machine learning techniques trained on human-provided emotional responses, ensuring alignment between music characteristics and user experience.”
Feature Comparison Table
| Key Feature of Claim 1 | US20220245193A1 | US20200228596A1 | US20150268800A1 | EP2843860A1 | WO2010027509A1 |
| Selecting a song based on computer-derived comparison between song representation and known similarities in other song representations | Fully Disclosed | Fully Disclosed | Partially Disclosed | Partially Disclosed | Partially Disclosed |
| Known similarities in song representations are based on a human-trained machine | Fully Disclosed | Partially Disclosed | Not Disclosed | Not Disclosed | Fully Disclosed |
| Song representation is based on isolating and identifying frequency characteristics | Fully Disclosed | Fully Disclosed | Not Disclosed | Partially Disclosed | Fully Disclosed |
| Representations of other songs are based on human listening to isolate and identify frequency characteristics | Fully Disclosed | Partially Disclosed | Not Disclosed | Not Disclosed | Partially Disclosed |
| Frequency characteristics correspond to moods of the songs | Fully Disclosed | Fully Disclosed | Partially Disclosed | Partially Disclosed | Fully Disclosed |
| Selection is based on similarity between moods of the song and moods of other songs | Fully Disclosed | Fully Disclosed | Fully Disclosed | Partially Disclosed | Fully Disclosed |
How to Find Related Patents Using Global Patent Search?
Finding related patents is essential in assessing the validity and uniqueness of US11899713B2. The Global Patent Search (GPS) tool streamlines this process by offering:
Search by patent number or description – Instantly retrieve relevant patents related to music streaming, playlist generation, and song classification.
Source: GPS
Feature mapping for patent comparison – GPS identifies technical similarities between patents, helping to determine if an existing related patent challenges a claim.
Comprehensive matching results – Users can browse through a curated list of related patents to uncover overlapping features.
Detailed reports for legal analysis – GPS provides structured feature mapping reports, showcasing how each related patent aligns with the original patent’s claims.
Data-driven insights for stronger cases – Whether you are challenging a patent’s validity or defending against infringement claims, GPS delivers precise, data-backed results.
By leveraging Global Patent Search, companies can efficiently locate prior patents, analyze feature disclosures, and make informed legal decisions in patent disputes.
Don’t Let Unchecked Patents Cost You; Get the Competitive Edge Now!
Patent disputes are high-stakes battles, and US11899713B2 is no exception. Whether you are defending your innovation or challenging a questionable patent, you need hard-hitting data, not guesswork.
Global Patent Search (GPS) gives you the firepower to:
- Expose weak claims – Find related patents that could invalidate key features.
- Fortify your legal strategy – Leverage feature mapping to pinpoint critical overlaps.
- Dominate patent litigation – Arm yourself with bulletproof prior research before stepping into the courtroom.
Don’t play defense, take control. Start your search with Global Patent Search today!
Disclaimer: The information provided in this article is for informational purposes only and should not be considered legal advice. The related patent references mentioned are preliminary results from the Global Patent Search (GPS) tool and do not guarantee legal significance. For a comprehensive related patent analysis, we recommend conducting a detailed search using GPS or consulting a patent attorney.
