Machine Learning for Audio, AI based Noise Reduction

As the Voice User Interface(Voice UI) keep growing in different market segments but use of these systems in automotive and other portable devices will be challenging due to very loud surrounding noise like road noise, engine noise, wind noise, babble noise and so on. It is also challenging to enable the voice interface for the products where we have more than stereo play back like Automotive Audio and TV Soundbars. Adopting Acoustic Echo Cancellation (AEC) per speaker will overload the DSP and SoCs.
Listen Noisy Audio & Noise Reduction Samples
For Noise Reduction Click On/Off button in Player
Baby Cry Noise Reduction
KeyBoard Noise Reduction
Dog Barking Noise Reduction
Pressure Cooker Noise Reduction
Wind Noise Reduction

AI based NOISE REDUCTION

AMI Technologies come up with the machine learning techniques to remove the surround noise and all other interfaces noise by which it improves the voice quality drastially for Voice UI devices, Smart Phones/Tabs/Laptops and Video recoding devices so on. Our machine learning noise reduction keeps learning from the environment and adapts to it quickly and removes the surround and interface noise within no time. As a result OEMs and ODMs are accelerating their developments to enable the Voice Control for their products.

Noise Reduction for Automotive applications:

Hands free communication is very vital for driver safety but due to the very noisy surround like road noise, wind noise, audio playback in the car voice calls and voice control is very challenging. AMI Technologies machine learning noise reduction algorithms quickly adapts to the surroundings and improves the voice quality by which we can do truly hands free communications in Automotive. We can control any device with the voice or have a crystal clear Voice or Video calls on the Go!.

Noise Reduction for Video Cameras and Mobile devices:

While today’s Video cameras capture 3D and 4K Video but most of the time audio is very noisy due to the surround noises and wind noise when we are at home or on roads. Even after our smart devices can power so many apps but making the voice calls or video calls when we are in subway train station or airport or in busy restaurant is very difficult and it is still challenging to make clear call in these enviroments. AMI Technologies Noise Reduction algorithms removes these noises and improves the voice quality drastically. So by enabling the AMI Technologies Noise Reduction for video cameras and mobile phones we can have crystal clear voice.

Tuning Tool

  1. Flexibility to real time tuning and further tweaking the signal flow is important while building the product, OEMs and ODMs also want to differentiate their product with competitors by scale the system with new features. AMI Technologies provides its rapid prototyping system to fine tune the design and re design the signal flow to meet above needs. It is 5 stage development Tool where user can design graphically Generate Code build & Run Tune on Fly automate the Testing Ship the product

  2. It has the rich set of the inspectors to tune the system

  3. Field proven and well defined host communication and tuning protocols over Serial peripheral interface(SPI), two wire interface (I2C) , UART and Ethernet / WiFi

Platforms Supported

  1. ADI SHARC and Blackfin Processors
  2. ARM Cortex-M7/M4 Processors
  3. ARM Cortex-A based SOCs (Linux and Android OS)
  4. Qualcomm Snapdragon Hexagon DSP
  5. Qualcomm Kalimba DSP
  6. Low cost DSPs (contact for list)
  7. Intel x86 Devices

Application Areas:

  1. Voice UI systems /Personal Assistance systems
  2. Automotive
  3. Video Cameras
  4. Smart phones/Tabs/Laptops
  5. Gaming Headsets
  6. Drones

For more details please contact us at contact@amitekh.com