US 12125496

I was unable to locate a US patent with the exact number 12125496 in the available databases. A search for "US12125496B1" on Google Patents returned a message indicating that the patent number could not be found.

Therefore, I cannot provide the title, assignee, inventors, filing date, issue date, abstract, or a plain-language overview of the independent claims for US patent 12125496.

Regarding the CAFC 2026 dockets, searches did not reveal any cases specifically pertaining to a patent numbered 12125496. The provided search results for CAFC dockets offer general information about scheduled cases but do not list any litigation for this specific patent number.

As of April 26, 2026, the following litigation is known involving US patent 12125496:

• Case: US case filed in California Northern District Court [cite: 3:25-cv-05666]

  • Plaintiff(s): Not specified in the provided information.
  • Defendant(s): Not specified in the provided information.
  • Jurisdiction: California Northern District Court [cite: 3:25-cv-05666]
  • Case Number: 3:25-cv-05666 [cite: 3:25-cv-05666]
  • Filing Date: Not specified in the provided information.
  • Outcome or Current Status: Active [cite: 3:25-cv-05666]

• Case: PTAB case IPR2026-00273 filed (Pending) [cite: IPR2026-00273]

  • Plaintiff(s): Petitioner (not specified by name in the provided information) [cite: IPR2026-00273]
  • Defendant(s): Not specified in the provided information.
  • Jurisdiction: Patent Trial and Appeal Board (PTAB) [cite: IPR2026-00273]
  • Case Number: IPR2026-00273 [cite: IPR2026-00273]
  • Filing Date: Not specified in the provided information.
  • Outcome or Current Status: Pending [cite: IPR2026-00273]

Proceedings overview

One Inter Partes Review (IPR) proceeding, IPR2026-00273, has been filed against US patent 12125496 and is currently pending. This means the patent's claims are actively being challenged, and the defensive posture for a defendant facing assertion is uncertain, awaiting the outcome of this IPR.

IPR2026-00273 — Krisp Technologies, Inc. v. Sanas Ai Inc

• Type: Inter Partes Review
• Filed: 2026-03-26
• Status: Pending. This proceeding is in the initial stages and has not yet reached an institution decision.
• Judge panel: Not yet publicly available as an institution decision has not been issued.
• Petition grounds: Details regarding specific claims challenged, prior art references, and statutory bases (§ 102 / § 103 / § 112) are not yet publicly available in a decision.
• Institution decision: Not yet issued. The statutory deadline for an institution decision is typically six months from the filing date, which would be approximately 2026-09-26.
• Final Written Decision: Not applicable; no FWD has been issued.
• Settlement / termination: Not applicable; the proceeding is active.
• Appeal: Not applicable; no FWD has been issued to appeal.
• Defensive value: This active IPR proceeding indicates that the validity of US12125496 is currently under scrutiny by Krisp Technologies, Inc. A defendant facing assertion of this patent should monitor this IPR closely, as an institution decision and subsequent FWD could significantly impact the strength of the patent. If the IPR is instituted and claims are eventually invalidated, it could substantially weaken the patent owner's position.

Strategic summary

Currently, all claims of US patent 12125496 are untested by a final PTAB decision. IPR2026-00273, filed by Krisp Technologies, Inc., is in its early stages and has not yet reached an institution decision. Therefore, no claims have been canceled or sustained through a PTAB trial or subsequent appeal. The patent is currently facing a live challenge, indicating potential weaknesses that Krisp Technologies, Inc. believes it can demonstrate.

The estoppel landscape under § 315(e)(2) will only become relevant for Krisp Technologies, Inc. (and their privies) if the IPR proceeds to a Final Written Decision. At this point, all prior-art grounds are theoretically available to other potential defendants. The fact that Krisp Technologies, Inc., a company also involved in voice enhancement technology (as suggested by their prior art citations in other patents), has filed an IPR could signal a competitive interest in challenging this patent.

Recommended next steps

• As IPR2026-00273 is pending, the most critical upcoming milestone is the institution decision. This decision is expected around 2026-09-26. A defendant should actively monitor the USPTO PTAB E2E system for updates on this proceeding.
• If the IPR is instituted, the trial will commence, leading to an oral hearing and ultimately a Final Written Decision, typically within one year of institution. These dates will become available if the trial is instituted.
• The absence of any concluded PTAB activity means that the claims of US12125496 remain presumed valid and untested by the PTAB. This also means that for other potential challengers, the full range of prior art arguments remains available for new IPR petitions, should they choose to file one.

I am unable to perform live searches on the USPTO Patent Assignment Search database at assignmentcenter.uspto.gov or assignment.uspto.gov/patent/index.html. Therefore, I cannot reconstruct the full assignment record for US patent 12125496 to identify any NPE/patent-troll patterns.

My previous search for "US12125496B1" on Google Patents also indicated that the patent number could not be found, which suggests there might be an issue with the patent number provided or its indexing in public databases at the time of that prior check. However, based on the full patent text provided in the prompt, which is authoritative, I can extract some information.

Since I cannot access the USPTO assignment database, I can only provide information based on the provided patent text.

Inventors

• Shawn Zhang (Sanas Ai Inc.)
• Lukas Pfeifenberger (Sanas Ai Inc.)
• Jason Wu (Sanas Ai Inc.)
• Piotr Dura (Sanas Ai Inc.)
• David Braude (Sanas Ai Inc.)
• Bajibabu Bollepalli (Sanas Ai Inc.)
• Alvaro Escudero (Sanas Ai Inc.)
• Gokce Keskin (Sanas Ai Inc.)
• Ankita Jha (Sanas Ai Inc.)
• Maxim Serebryakov (Sanas Ai Inc.)

All named inventors are associated with Sanas Ai Inc., the original assignee, at the time of the patent's filing.

Original assignee

The original assignee named on the issued patent is Sanas Ai Inc.

Based on the patent description, Sanas Ai Inc. develops "methods and systems for voice enhancement using neural networks" which are used in applications like customer service, medical fields, and education. The patent describes a "voice enhancement system 100" which suggests they ship a product embodying the claims.

The current status of Sanas Ai Inc. is "Active" as per the legal status information provided.

Assignment timeline

The Google Patents legal status section for US12125496B1 indicates an assignment event:

• 2024-07-11 — Assigned to Sanas.ai Inc.
  • Conveyance: ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS).
  • Assignors: PFEIFENBERGER, Lukas, Dura, Piotr, Braude, David, BOLLEPALLI, BAJIBABU, WU, Jilong, Escudero, Alvaro, Jha, Ankita, KESKIN, GOKCE, SEREBRYAKOV, Maxim, ZHANG, SHAWN
  • Assignee: Sanas.ai Inc.
  • Context: This appears to be an internal re-assignment or confirmation of assignment from the inventors to the company, Sanas.ai Inc., after the application was filed by "Sanas Ai Inc" and before it was granted. This is a common practice to formalize ownership.

As I cannot access the live USPTO Assignment Center, I am unable to provide reel/frame numbers or correspondent information for this event, nor can I search for any other assignments.

Timeline diagram

timeline
    title Ownership of US 12125496
    2023 : Priority date
    2024 : Filed by Sanas Ai Inc
    2024 : Inventors assigned to Sanas.ai Inc
    2024 : Application granted
    2024 : Published

NPE / troll-pattern signals

Due to my inability to access the USPTO Assignment Center and the lack of comprehensive assignment records in the provided patent text, I cannot confidently assess most of these signals.

1. Shell-entity transfer — Unclear. Without USPTO assignment records, I cannot determine if there have been transfers to shell entities. The Google Patents entry only shows an assignment from inventors to Sanas.ai Inc.. Sanas Ai Inc. appears to be an operating company based on its description of products and services.
2. Known asserter in the chain — Not present. The only assignee identified is Sanas Ai Inc., which is not a known NPE from public lists.
3. Repeat correspondent across the chain — Unclear. I cannot determine this without access to USPTO assignment records which list the correspondent of record.
4. Cascading transfers — Not present. Only one assignment event is noted, from inventors to the company.
5. Pre-litigation transfer — Unclear. While there is litigation noted, I cannot determine if any assignment was made within 6 months of the first suit without access to the full litigation timeline and assignment records.
6. Bankruptcy fire-sale — Not present. The legal status of Sanas Ai Inc. is "Active", and there is no indication of bankruptcy.
7. Privateering — Unclear. There is no information in the provided patent text or search results to suggest privateering.
8. Defensive aggregator (anti-NPE) — Not present. The patent is currently assigned to Sanas Ai Inc., not a defensive aggregator.

Verdict

Insufficient data.

The verdict is "Insufficient data" because I could not access the USPTO Assignment Center to retrieve the full assignment history, including reel/frame numbers and correspondent information. The available information from Google Patents only indicates a single assignment from the inventors to the original assignee, Sanas.ai Inc. on 2024-07-11. This limited data prevents a comprehensive analysis of NPE/troll patterns.

I must first address the discrepancy regarding US patent 12125496. The previously generated section stated an inability to locate US patent 12125496. However, the comprehensive patent text provided for this task, titled "US12125496B1 - Methods for neural network-based voice enhancement and systems thereof," confirms its existence and provides all relevant details. Therefore, I will proceed with the analysis of US12125496B1 based on the provided authoritative patent text.

The USPTO website provides tools for searching patents, such as Patent Public Search. While I can perform a search, the provided patent text for US12125496B1 already contains the "Citations" section, which lists the prior art identified by the examiner. For the purpose of this analysis, the cited patents within US12125496B1 are considered the most relevant prior art.

US Patent 12125496B1 Details

• Title: Methods for neural network-based voice enhancement and systems thereof
• Publication Number: US12125496B1
• Assignee: Sanas Ai Inc
• Inventors: Shawn Zhang, Lukas PFEIFENBERGER, Jason Wu, Piotr Dura, David Braude, Bajibabu Bollepalli, Alvaro Escudero, Gokce Keskin, Ankita Jha, Maxim Serebryakov
• Filing Date: 2024-04-24
• Issue Date (Publication Date): 2024-10-22
• Abstract: The disclosed technology relates to methods, voice enhancement systems, and non-transitory computer readable media for real-time voice enhancement. In some examples, input audio data including foreground speech content, non-content elements, and speech characteristics is fragmented into input speech frames. The input speech frames are converted to low-dimensional representations of the input speech frames. One or more of the fragmentation or the conversion is based on an application of a first trained neural network to the input audio data. The low-dimensional representations of the input speech frames omit one or more of the non-content elements. A second trained neural network is applied to the low-dimensional representations of the input speech frames to generate target speech frames. The target speech frames are combined to generate output audio data. The output audio data further includes one or more portions of the foreground speech content and one or more of the speech characteristics.

Most Relevant Prior Art for US12125496B1

The following patent citations were identified as prior art by the examiner for US12125496B1:

1. US11410684B1

   • Full Citation: US11410684B1: Text-to-speech (TTS) processing with transfer of vocal characteristics, assigned to Amazon Technologies, Inc.
   • Publication Date: 2022-08-09
   • Priority Date: 2019-06-04
   • Brief Description: This patent generally relates to text-to-speech (TTS) processing and, more specifically, to methods for transferring vocal characteristics. While not directly focused on enhancing degraded speech, it deals with manipulating and transferring vocal characteristics, such as voice identity, which is a speech characteristic that US12125496B1 aims to preserve and enhance.
   • Potential Anticipation of Claims under 35 U.S.C. § 102: This patent may potentially anticipate aspects of claims 1, 11, and 16 that involve handling or preserving "speech characteristics" (e.g., voice identity). However, its primary focus on Text-to-Speech (TTS) rather than real-time voice enhancement from noisy input, and its lack of the specific two-neural-network architecture for dimensionality reduction and non-content element omission, suggests it does not anticipate the core invention. A detailed comparison would require examining the specific methods used for characteristic transfer in US11410684B1.

2. US11482235B2

   • Full Citation: US11482235B2: Speech enhancement method and system, assigned to Qnap Systems, Inc.
   • Publication Date: 2022-10-25
   • Priority Date: 2019-04-01
   • Brief Description: This patent broadly covers a speech enhancement method and system. The title directly aligns with the subject matter of US12125496B1, indicating a focus on improving the quality and clarity of speech signals.
   • Potential Anticipation of Claims under 35 U.S.C. § 102: Given its general title, US11482235B2 is highly likely to be considered relevant prior art for claims 1, 11, and 16 of US12125496B1, especially regarding the broad concepts of "voice enhancement system," "method for real-time voice enhancement," and "non-transitory computer-readable medium comprising instructions... to enhance speech." Without reviewing the detailed claims and description of US11482235B2, it is difficult to determine if it discloses the specific two-neural-network architecture, the conversion to low-dimensional representations that omit non-content elements, and the dynamic generation of target speech frames, which are key distinctions of US12125496B1. However, it certainly covers the broad field.

3. US11705147B2

   • Full Citation: US11705147B2: Mixed adaptive and fixed coefficient neural networks for speech enhancement, assigned to Qualcomm Incorporated.
   • Publication Date: 2023-07-18
   • Priority Date: 2020-04-29
   • Brief Description: This patent describes the use of "mixed adaptive and fixed coefficient neural networks for speech enhancement." This directly addresses the application of neural networks for speech enhancement, a core aspect of US12125496B1.
   • Potential Anticipation of Claims under 35 U.S.C. § 102: US11705147B2 is highly relevant as it explicitly uses neural networks for speech enhancement, which is the foundational technology for US12125496B1's claims 1, 11, and 16. It could potentially anticipate the use of neural networks for speech enhancement in general. The specifics of "mixed adaptive and fixed coefficient" might represent a different architectural approach compared to US12125496B1's two-network system for dimensionality reduction and reconstruction while preserving speech characteristics and omitting non-content elements. A detailed analysis would compare whether US11705147B2's methods inherently or explicitly disclose the two-stage NN process with low-dimensional representation that selectively omits non-content elements.

4. US11868883B1

   • Full Citation: US11868883B1: Intelligent control with hierarchical stacked neural networks, assigned to Michael Lamport Commons.
   • Publication Date: 2024-01-09
   • Priority Date: 2010-10-26
   • Brief Description: This patent focuses on "intelligent control with hierarchical stacked neural networks." This describes a general neural network architecture, specifically hierarchical and stacked networks, within the context of intelligent control.
   • Potential Anticipation of Claims under 35 U.S.C. § 102: This patent might be relevant for its general disclosure of "hierarchical stacked neural networks." US12125496B1's detailed description mentions that the low-dimensional representation "may be achieved by using a hierarchical feature extraction network" (though this is not explicitly in the independent claims). Therefore, if a claim of US12125496B1 were to broadly cover any hierarchical neural network, US11868883B1 could be anticipatory. However, as it stands, the claims of US12125496B1 define a specific application of two neural networks for voice enhancement, focusing on noise reduction through low-dimensional representation, which is distinct from the general "intelligent control" context of US11868883B1. It is less likely to anticipate the specific voice enhancement method and system claimed in US12125496B1 than the patents more directly related to speech processing.

Correction Regarding Patent 12125496

It must be noted that the previous analysis incorrectly stated that US patent 12125496 could not be located. The full patent text provided for US12125496B1 is authoritative for this analysis, and all subsequent analysis will proceed using this document as the ground truth.

Obviousness Analysis of US12125496B1 Under 35 U.S.C. § 103

Patent Identification: US12125496B1, titled "Methods for neural network-based voice enhancement and systems thereof."
Priority Date: 2023-05-05.
Current Date: April 26, 2026.

Level of Ordinary Skill in the Art (POSITA)

A person having ordinary skill in the art (POSITA) relevant to US12125496B1 would be an individual with a strong background in digital signal processing, machine learning (especially deep learning), and audio/speech processing. This would typically entail a Master's degree or equivalent professional experience in fields such as Electrical Engineering, Computer Science, or Artificial Intelligence, with expertise in neural network architectures, audio feature extraction, and speech enhancement techniques.

Overview of Independent Claims

The independent claims of US12125496B1 describe a two-stage neural network approach for real-time voice enhancement:

• Claim 1 (System): A voice enhancement system that fragments input audio data (containing foreground speech content, non-content elements, and speech characteristics) into speech frames. A first neural network converts these frames into low-dimensional representations, explicitly omitting one or more non-content elements. A second neural network then applies to these low-dimensional representations to generate target speech frames, which are combined to form output audio data retaining foreground speech content and speech characteristics.
• Claim 11 (Method): A method involving training both the first and second neural networks. The trained first neural network converts input speech frames to low-dimensional representations, omitting non-content elements. The trained second neural network applies to these low-dimensional representations to generate target speech frames, which are combined to produce output audio data comprising foreground speech content and speech characteristics.
• Claim 16 (Non-transitory computer-readable medium): Instructions for digitizing input audio, fragmenting it, converting frames to low-dimensional representations (omitting non-content elements) via a first neural network, applying a second neural network to generate target speech frames, combining them into output audio data, and converting to analog output.

A core innovative aspect highlighted is the generation of a "low-dimensional representation" that omits non-content elements while preserving speech characteristics, acting as an intermediate step between two neural networks.

Cited Prior Art References

The following prior art references, with priority dates preceding US12125496B1's priority date (2023-05-05), were identified:

1. US11410684B1 (Amazon Technologies, Inc.): Priority Date: 2019-06-04. Relates to Text-to-Speech (TTS) processing with transfer of vocal characteristics.
2. US11482235B2 (Qnap Systems, Inc.): Priority Date: 2019-04-01. Discloses a speech enhancement method and system using a deep neural network to predict a target spectrum diagram from an audio signal's spectrum.
3. US11705147B2 (Qualcomm Incorporated): Priority Date: 2020-04-29. Describes methods, systems, and devices for speech enhancement using mixed adaptive and fixed coefficient neural networks.
4. US11868883B1 (Michael Lamport Commons): Priority Date: 2010-10-26. Concerns an intelligent control system with hierarchical stacked neural networks.

Obviousness Analysis (35 U.S.C. § 103)

The primary problem addressed by US12125496B1, as stated in its background, is that existing voice enhancement techniques, while capable of reducing noise, often distort essential speech features, leading to inaccurate automatic speech recognition (ASR) results or failing to improve naturally unclear speech. This identifies a long-felt need and a known problem in the art.

The independent claims of US12125496B1 introduce a two-neural-network architecture where a first network creates a low-dimensional representation of input speech frames that specifically omits non-content elements (e.g., background noise, microphone pops, low-fidelity audio) but retains foreground speech content and characteristics, before a second network reconstructs enhanced target speech.

While no single prior art reference explicitly discloses this exact two-stage, noise-omitting low-dimensional representation, a POSITA would have been motivated to combine existing knowledge and techniques from the cited prior art and the general state of the art to arrive at the claimed invention.

Combination of Prior Art and Motivation:

A POSITA, in 2023, would be aware of the following:

• Neural Network-Based Speech Enhancement: Both US11482235B2 and US11705147B2 teach the use of neural networks for speech enhancement. US11482235B2 uses a deep neural network to predict a target spectrum diagram for generating an enhanced audio signal, and US11705147B2 uses mixed adaptive and fixed coefficient neural networks for general speech enhancement. These patents would provide the basic teaching of employing neural networks for enhancing speech signals.
• Hierarchical/Stacked Neural Networks: US11868883B1 demonstrates the concept of hierarchical stacked neural networks for complex tasks like intelligent control. While not directly in speech enhancement, it teaches that multi-stage or cascaded neural network architectures are a known and effective approach for breaking down and solving complex problems by processing information through successive layers or networks. A POSITA would readily recognize the applicability of a multi-stage neural network architecture to the complex problem of robust speech enhancement.
• Feature Extraction and Dimensionality Reduction: The specification of US12125496B1 itself describes that the low-dimensional representation can be achieved by "pre-processing the input audio data 402 to remove noise and other distortions" using "a noise reduction algorithm... or a filtering technique" and that "features may be extracted... such as by using Fourier Transform, Mel-Frequency Cepstral Coefficients (MFCC)... encoded... using techniques such as Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), or other dimensionality reduction techniques." These are well-known signal processing and machine learning techniques for isolating and compressing relevant information while reducing noise or irrelevant data.

Motivation for Combination:

Given the acknowledged problem that existing single-stage neural network enhancement methods (e.g., as taught by US11482235B2 or US11705147B2) often distort desired speech characteristics while attempting to remove noise, a POSITA would be strongly motivated to improve upon these systems. The motivation would be to achieve more precise noise suppression without compromising speech intelligibility and integrity.

To address this, a POSITA would consider:

1. Breaking down the problem: Instead of a single network trying to do everything, separating the task into distinct stages using a multi-network approach (informed by US11868883B1).
2. Explicitly isolating speech content: Leveraging known feature extraction (e.g., MFCC) and dimensionality reduction techniques (e.g., PCA, LDA) to create an intermediate representation that focuses only on the critical speech characteristics and filters out or minimizes the non-content elements (noise, pops, etc.). This step directly addresses the problem of distortion by ensuring that the signal presented for enhancement is already stripped of undesirable components.
3. Reconstruction: Using a second network to reconstruct the full speech signal from this cleaner, low-dimensional representation, ensuring that the output retains the desired speech characteristics but is free from the previously omitted non-content elements.

The combination of:

• Neural network-based speech enhancement (e.g., US11482235B2 or US11705147B2)
• The architectural concept of multi-stage/hierarchical neural networks (e.g., US11868883B1)
• Well-known techniques for feature extraction, noise reduction, and dimensionality reduction (as described in the background of US12125496B1)

would lead a POSITA to develop the claimed two-network system with an intermediate low-dimensional, noise-free representation. The motivation is to overcome the known trade-off between noise reduction and speech distortion in prior art systems by explicitly separating and processing speech content from non-content elements. This approach represents an expected engineering choice to improve performance in a known problem area.

Therefore, the methods and systems described in independent claims 1, 11, and 16 of US12125496B1 would be obvious to a POSITA by combining the teachings of the cited prior art and general knowledge in the field to address a recognized technical challenge.