Meta Platforms, Inc. has secured Swanson Reed’s Patent of the Month for February 2026 in the Film, Media and Telecommunications industry for an outstanding invention. This innovation focuses on a newly recognized patent titled ‘Systems and methods for securely captioning video calls’. The patent describes a computer-implemented method for securely captioning video calls by parsing speech directly on a client device and transmitting the transcript to an additional device, bypassing the need for cloud-based processing.
Revolutionizing Secure Communications
Abstract: A computer-implemented method for securely captioning video calls may include (i) detecting, by a messenger application executing on a client device, speech captured by a microphone of the client device and video of a speaker of the speech being captured by a camera of the client device, (ii) parsing, by the messenger application on the client device, the speech to create a transcript of the speech, and (iii) transmitting, by the messenger application on the client device, the transcript of the speech to an additional device for display to a user of the messenger application in combination with the video of the speaker. Various other methods, systems, and computer-readable media are also disclosed.
U.S. R&D Tax Credit Eligibility
The development of the technology outlined in this patent is an excellent example of work that meets the IRS four-part test for the U.S. Research and Development (R&D) Tax Credit:
- Permitted Purpose: The development aims to improve software performance, reliability, and security by creating a new method to process speech-to-text locally on a device, ensuring end-to-end encryption for video call captions.
- Technological in Nature: The innovation relies entirely on the hard sciences of computer science and software engineering, specifically utilizing natural language processing (NLP), audio/video synchronization, and secure network transmission protocols.
- Elimination of Uncertainty: At the onset of development, there was technical uncertainty regarding how to efficiently run complex transcription algorithms directly on mobile processors without causing severe battery drain, latency, or device overheating.
- Process of Experimentation: Software engineers had to systematically test multiple algorithmic models, evaluate various data compression techniques for sending the transcript alongside the video, and iterate on different architectural designs to achieve real-time synchronization.
3 Practical Applications Qualifying for R&D Tax Credits
- Developing Enterprise-Grade, HIPAA-Compliant Messengers: Designing a proprietary corporate video tool that requires local on-device captioning to comply with strict medical or legal data privacy standards. The software engineering and experimentation involved in optimizing accurate NLP models to run entirely offline on a smartphone without cloud assistance directly qualifies for R&D credits.
- Low-Bandwidth Accessibility Engineering: Creating a specialized video call mode tailored for rural or developing regions with poor internet infrastructure. The R&D involves engineering the software to perfectly synchronize locally parsed text transcripts with heavily compressed, low-bitrate video streams, requiring extensive testing to overcome latency and packet-loss uncertainties.
- Real-Time Multilingual On-Device Translation: Expanding this patented system to not only transcribe but actively translate speech directly on the client device before securely transmitting it to the receiving party. The highly complex engineering required to run dual-NLP models (transcription and translation) efficiently on consumer mobile hardware involves overcoming significant technical hurdles and iterating through multiple software designs.
