Lingxi Tech — On-Device AI Audio

Software / Algorithm Provider

Full-stack on-device AI audio: noise reduction, echo cancellation, beamforming, source localization, wake-up, offline recognition, voiceprint and acoustic detection

Platform verifiedCompany claimed

Platform verification does not guarantee product quality or fulfillment.

Lingxi Tech — On-Device AI Audio

Contact manufacturer

Phone 19928743372
Address Shenzhen: Rm 1601A, Bldg B, Fencheng Zhigu, Xixiang Subdistrict, Baoan District, Shenzhen, Guangdong; Shanghai: Rm 1M, Bldg 1, No.95, Lane 85 Cailun Rd, Pudong New Area, Shanghai
Certifications Joint acoustic lab and incubation platform with Zhejiang Tsinghua Yangtze Delta Institute; core team from Fudan microelectronics, with AMD/NXP chip experience.

Overview

Lingxi Tech, founded early 2021 in Zhangjiang, Shanghai, is an AI company focused on on-device audio algorithms. For smart hardware, industrial motors, e-bikes, automotive and robots, it provides full-stack voice enhancement, voiceprint detection, voice interaction and acoustic sensing so devices hear clearly, understand and recognize locally.

The algorithm stack has three layers: audio front-end (noise reduction, echo cancellation, beamforming, source localization), interaction (wake-up, offline recognition, voiceprint reduction, detection models) and audio separation (music/voice). It runs offline on device with low power and high privacy, supporting custom wake words and fast model iteration. Delivery covers SDK, private deployment and custom algorithm development.

Solutions Offered

Main products

  • Audio front-end: noise reduction, echo cancellation, beamforming pickup, sound-source localization.
  • Voice interaction: wake-up, offline ASR, voiceprint denoise, acoustic detection models.
  • Audio separation: music / vocal separation.
  • Delivery: acoustic algorithm SDK, private deployment, custom algorithm development.
  • Scenarios: smart hardware, motor acoustic QC, e-bikes, humanoid / service / cleaning robots, pet and special-sound detection.

Comments

No comments yet. Be the first to comment.