User:Harry 0405/sandbox
| Industry | AI Audio Technology |
|---|---|
| Founded | 2015 |
| Founder | Henney Oh, Ted Lee |
| Headquarters | 2F, 505, Teheran-ro, Gangnam-gu, Seoul, Republic of Korea / 2120 University Ave, Berkeley, CA 94704, USA |
| Website | https://www.gaudiolab.com/ |
Gaudio Lab is a South Korean AI audio technology startup, established on May 6, 2015. The company specializes in advanced AI audio solutions, including AI source separation technology and AI lyric synchronization technology. It has garnered international recognition, winning the Innovation Award at CES, the largest ICT convergence expo globally, and reaching the finalist stage in the audio experience category at the SXSW Innovation Awards, part of a major music/film/media festival in the United States.[1] Gaudio Lab's technologies have been endorsed as standards by global standardization bodies such as MPEG and 3GPP.[2] Gaudio Lab has 9 Ph.D. holders and 6 M.S. in acoustic engineering. This indicates that the company employs a professional sound engineering workforce that exceeds the size of those found in many large corporations, including Samsung Electronics.
Key Technologies[edit]
Source Separation[edit]
This technology enables the extraction of individual sound sources from mixed audio signals that contain multiple overlapping sounds. This technology is used in various contexts like separating instruments (stem separation), extracting voices, getting rid of vocals, reducing noise, and making mixed reality (MR) experiences. Gaudio Lab's source separation technology boasts a high separation efficiency, achieving a Signal-to-Distortion Ratio (SDR) of 10.02. It is considered superior to similar technologies from major companies like Sony, Meta, and Deezer.[3]
Spatial Audio[edit]
Spatial Audio is an audio technology that identifies the location of sounds and delivers them to the user's ears in three dimensions via binaural rendering. Gaudio Lab's spatial audio technology has been adopted as the MPEC standard.[4] It can be implemented across various environments, including platforms, true wireless stereo (TWS) systems, and live streaming services.
Loudness Normalization (LM1: Loudness Management 1)[edit]
Loudness Normalization technology standardizes the volume levels across all content on various platform, ensuring an optimal playback for users. Gaudio Lab's loudness normalization technology is utilized in a range of applications including music/VOD streaming services, televisions, tablets, and mobile devices, etc. It has been established as a national standard by the Korea Telecommunication Technology Association (TTA) and also as an international standard by the Consumer Technology Association (CTA).[5]
Generative Audio AI (FALL-E)[edit]
FALL-E is a generative sound AI technology designed to automatically create suitable sounds that complement specific content. Gaudio Lab’s FALL-E drastically reduces the processing time required for foley and sound effect creation, by more than 80%.[6] To train FALL-E, 10,000+ hours of high-quality data were utilized. In March 2024, content featuring audio created by FALL-E was published, layered over video produced by OpenAI’s video generative AI, Sora. FALL-E was also showcased to Microsoft’s CEO Satya Nadella during his visit to Gaudio Lab's CES booth in 2024, who remarked "amazing", after experiencing the technology.[7]
Main Products[edit]
Just Voice[8][edit]
Just Voice is a real-time noise suppression solution that effectively removes background noises, delivering clear voice. It operates with ultra-low latency of 30ms and requires minimal computational resources, making it suitable for embedded systems. Just Voice can be applied on various operating systems, including Android, iOS, Windows, macOS, and Linux. The product earned the CES 2024 Innovation Award and was a finalist in the SXSW 2024 Innovation Awards under the Audio Experience category. Gaudio Lab released a macOS application called Just Voice Lite, which also features real-time noise cancellation.
AI Text Sync (GTS)[9][edit]
GTS represents Gaudio Lab's real-time lyric synchronization technology, where AI automatically aligns lyrics with sound sources swiftly and accurately. Unlike manual synchronization, which typically requires 240 seconds to sync a four-minute song, GTS accomplishes this in just 5 seconds. It also provides detailed syncing capabilities, including line-by-line and word-by-word adjustments. Currently, the technology supports Korean, English, Chinese, and Japanese languages.
LM1(Loudness Management Solution)[10][edit]
It is a loudness solution ready for OTT/music streaming. It reduces the loudness variance between sources to protect users’ hearing. Unlike in typical file-based solutions, the metadata-driven “server-client ” architecture eliminates the need for additional transcoding. It can set different target loudness for TV and mobile for one source saved in the server. It has been recongnized by Consumer Technology Association (CTA) and American National Standards Institute (ANSI) as an international standard. It won CES Innovation Award 2023 Honoree for streaming.
LM1 is a loudness solution designed for OTT and music streaming services, aimed at minimizing volume discrepancies between different pieces of content. This solution operates on a metadata-driven 'server-client' architecture, which eliminates the need for additional transcoding typical of file-based solutions. It allows for distinct target loudness settings for TV and mobile devices using a single audio source stored on the server. LM1 has been recognized as an international standard, adopted by both the Consumer Technology Association (CTA) and the American National Standards Institute (ANSI). It has also received the CES Innovation Award.
GSP - SA[11][edit]
GSP-SA is a fully software-based solution designed to facilitate the creation of spatial audio content without the need for professional hardware. It is optimized for streaming, boasting an exceptionally low latency of 0.26 seconds. When integrated with a video switcher and live streaming software, it automatically processes and transmits audio that is suited to the visual content of the scene.
Gaudio Spatial Audio (GSA)[12][edit]
Gaudio Spatial Audio(GSA) is Gaudio Lab’s proprietary high-quality spatial audio solution, which operates with an low latency of less than 0.05 seconds. Utilizing binaural rendering technology, GSA incorporates over 10 registered patents to accurately simulate psychoacoustic effects. The technology is versatile, supporting various audio formats including stereo, immersive audio, and ambisonics.
Gaudio Studio[13][edit]
Gaudio Studio is a source separation tool powered by GSEP, Gaudio Lab's proprietary technology that isolates individual instruments, voices, and noises. It is utilized for vocal removal, instrument practice, and karaoke(noraebang) track production. Notably, Gaudio Studio was rated the highest among five instrument separation software tools in an evaluation conducted by Music Radar, an international music publication.[14]
Partners[edit]
- Tving: Korean #1 VOD streaming service. Utilizes Gaudio Lab’s sound equalization technology, LM1 (Loudness Management 1), to reduce volume differences between contents.
- NHN Bugs: Giant music streaming platform. Commercializes a real-time lyrics viewing service through AI Text Sync, which synchronizes songs and lyrics using Gaudio Lab’s AI Source Separation (GSEP) technology.
- VIBE: Giant music streaming platform owned by Naver Corp. Improves the synchronization accuracy of melody and lyrics using Gaudio Lab’s GTS technology.
- LG Electronics: Integrates the Spatial Upmix function into the LG Velvet smartphone.
- VinSmart: The ARIS smartphone by VinSmart is outfitted with the Spatial Upmix function to offer an immersive sound experience. Additionally, it corrects limitations of the smartphone's built-in speaker by analyzing content volume and enhancing output levels.
- Casa Batllo: In the Casa Batllo project in Barcelona, a virtual audio guide is provided, enabling visitors to experience the dynamic movement of the space.
- Naver Now: Features an Immersive Audio function and minimizes volume discrepancies across video content within the VOD platform.
- CGV: CGV's ScreenX theaters, which project images on three sides, are equipped with a Spatial Audio function to enhance the viewing experience.
- FLO: Features a function that consistently adjusts different volume levels for each sound source.
Prize[edit]
- 2014: Binaural Rendering technology adopted as the ISO/IEC MPEG-H international standard.
- 2017: Received the Innovation of the Year Award at the AMD Studios VR Awards.
- 2017: Awarded VR Innovation Company of the Year at the VR Awards.
- 2020: Loudness technology ‘Metadata for loudness operation of streaming services’ established as a TTA standard (TTAK.KO-07.0146).
- 2022: LM1 (Loudness Management 1) adopted as the CTA-2075.1 standard.
- 2023: Winner of two CES 2023 Innovation Awards.
- 2023: Received the CES 2024 Innovation Award.
- 2023: Awarded the Director Award (1st place) at the 2023 Global ICT Standard Conference by the National Institute of Information and Communications Planning and Evaluation.
History[15][edit]
- April 2014: Binaural Rendering technology was adopted as the ISO/IEC MPEG-H international standard.
- May 2015: Gaudio Lab Co., Ltd. was established in Korea.
- July 2015: Gaudio Lab attracted seed investment from Softbank Ventures Korea and Capstone Partners.
- September 2016: Secured Series A investment from Korea Investment Partners, LB Investment, Softbank Ventures, and Capstone Partners.
- November 2016: Established Gaudio Lab, Inc. in the United States (California).
- January 2017: Released Works 1.0, a VR audio authoring tool.
- August 2017: Released Sol VR360 SDK 1.0, a VR audio platform solution.
- September 2017: Won the Innovation of the Year Award at AMD Studios VR Awards.
- October 2017: Received the VR Innovation Company of the Year Award at the VR Awards.
- December 2017: Launched Craft 1.0, a VR audio plugin for game engines.
- January 2018: Participated in CES 2018 Eureka Park.
- March 2018: Released Loudness SDK 1.0.
- September 2018: Released EQ SDK 1.0.
- January 2019: Released Integrated Audio SDK 1.0; registered the GAUDIO trademark in the US.
- December 2020: Loudness technology (Metadata for loudness operation of streaming services) established as a TTA standard (TTAK.KO-07.0146).
- October 2021: Secured Series B investment from LB Investment, Capstone Partners, DS Asset Management, Samsung Venture Investment, and Naver.
- March 2022: Launched the alpha service of Gaudio Studio, an online source separation service.
- April 2022: Collaborated with Dingo on the project 'Sound in the lab' and 'Who Will Lead the Next Generation?'.
- May 2022: Signed an agreement to acquire Wave Lab, one of the three major sound post production studio.
- June 2022: LM1 (Loudness Management 1) adopted as the CTA-2075.1 standard.
- September 2022: Launched GSA (v3.6.0) featuring the world's lowest Motion-to-Sound (M2S) Latency.
- November 2022: Featured GSEP (sound source separation) technology on JTBC’s Hidden Singer episode featuring Hyun-sik Kim.
- January 2023: Won two CES 2023 Innovation Awards.
- November 2023: Received the CES 2024 Innovation Award
- November 2023: Won the ICT Planning and Evaluation Institute President's Award (1st prize) at Global ICT Standards Conference 2023.
- March 2024: AI noise cancellation solution, Just Voice, was selected as a finalist in the audio experience category at the SXSW 2024 Innovation Award.
External Links[edit]
- Gaudio Lab - Official Website
- Instagram - Official Instagram
- X(Twitter) - Official X(Twitter)
- Youtube - Official Youtube
- Discord - Official Discord
Articles[edit]
- Gaudio Lab's AI Noise Reduction Tech Nominated for a SXSW Innovation Award - filmfestivals.com
- AudioSourceRE Sets New Standard for Real-Time, AI-Powered Audio Separation - audioxpress
- [Inside Korean Startup Renaissance]① Gaudio Lab eyes top spot in AI sound generation - The Chosun Daily
- Gaudio Lab Unveils Key AI Audio Tech for Spatial Computing Era at MWC 2024 - Business Korea
- SXSW Picks Gaudio Lab’s AI Noise-Canceling Technology as Finalist for Innovation Award - Business Korea
- We tested 5 of the best stem separation software tools (and the best one was free) - Music Radar
- "In 5~10 years, most of the movie sound will be Gaudiolab's work." - AI Times
- Korean startup to rival Dolby in unconquered 3D audio market - The Korea Economic Daily
References[edit]
- ^ Herh, Michael (2024-01-31). "SXSW Picks Gaudio Lab's AI Noise-Canceling Technology as Finalist for Innovation Award". Businesskorea (in Korean). Retrieved 2024-04-22.
- ^ "Patents & Publications". Gaudio Lab. Retrieved 2024-04-17.
- ^ "Source Separation". Gaudio Lab. Retrieved 2024-04-17.
- ^ ""In 5~10 years, most of the movie sound will be Gaudiolab's work."". AI타임스 (in Korean). 2023-07-06. Retrieved 2024-04-22.
- ^ "Loudness & Sound Quality". Gaudio Lab. Retrieved 2024-04-22.
- ^ "Generative Sound AI: FALL-E". Gaudio Lab. Retrieved 2024-04-22.
- ^ Suk-yee, Jung (2024-02-20). "Gaudio Lab Unveils Key AI Audio Tech for Spatial Computing Era at MWC 2024". Businesskorea (in Korean). Retrieved 2024-04-23.
- ^ "Just Voice, Gaudio Lab". Gaudio Lab. Retrieved 2024-04-17.
- ^ "AI Text Sync (GTS) ,Gaudio Lab". Gaudio Lab. Retrieved 2024-04-17.
- ^ "Loudness Normalization, Gaudio Lab". Gaudio Lab. Retrieved 2024-04-17.
- ^ "GSA for Live Streaming, Gaudio Lab". Gaudio Lab. Retrieved 2024-04-17.
- ^ "GSA (GAUDIO Spatial Audio), Gaudio Lab". Gaudio Lab. Retrieved 2024-04-17.
- ^ "GSEP Music, Gaudio Lab". Gaudio Lab. Retrieved 2024-04-17.
- ^ Mullenpublished, Matt (2023-09-28). "We tested 5 of the best stem separation software tools (and the best one was free)". MusicRadar. Retrieved 2024-04-22.
- ^ "Milestone, Gaudio Lab". Gaudio Lab. Retrieved 2024-04-17.