Licensing

Introduction

Whether you are in the car, in a conference room or in your living room, you want to be able to communicate clearly without having to shout. Philips’ BeClear Speech Enhancement algorithms deliver clear voice signal even in the noisiest environments enabling you to have a natural conversation or use voice commands to interact with your devices

Based on over 20 years of experience in speech enhancement, Philips BeClear is a flexible solution that can be easily tailored to your specific application or product. In addition to the software license and application libraries, customers also receive assistance on implementing the software for their customer platform and tuning the algorithms to deliver the best performance on each device.


BeClear Speech Enhancement

• One flexible suite of algorithms used for simultaneous voice communication and speech recognition for voice control usage
• Optimized for commercially available MCU and DSP platforms
• Manufacturer tuning to specific hardware form factor
• Tunable to any speech recognition (SR) engines without insight into SR pre-processing
• Optimized per use case (smart phone, smart watch, TV and voice control for IoT)
• Speeds up integration for new use cases
• All use case extensions based on proven technology

Download request

To download the document, please provide the following information.

 
 
 

Mobile / Wearable

Our mobile devices let us stay connected wherever we are. Using our phones or wearables in noisy environments can still be a challenge. It can be difficult to understand the person calling and we either end up talking loudly over the background noise or having to find a quiet space to take the call.

We also increasingly use our mobile phones in speaker mode, held at arm’s length, for video calls or to let others join in on the call. And this will only increase with wearable devices such as smartwatches.

Due to their small form factor, wearable devices are also driving a growing demand for voice input and control. The devices need to be able to recognize commands reliably to be able to execute them without the user needed to repeat them several times.

Whether using voice control or making a call, the BeClear Speech Enhancement algorithm ensures clear, relaxed and effortless communication.

Smart Home

In the smart home, there is also a trend towards increased voice interaction and control of our devices. We will no longer need to search for the remote control, we can just tell the television to switch channels or play the latest episode of our favorite show.

Because of their centralized position in the Smart Home, this means that these devices need to be able to capture our voices at distances larger than 5m.

Using voice control, the BeClear Speech Enhancement algorithm ensures clear, relaxed and effortless interactions.

Automotive

We still spend a lot of time on the road. Hands-free calling enables us to stay in touch while in the car. There is also is a growing demand for using voice commands to operate the smart features in our car, from programming the GPS to making a call or changing the radio station. Voice control is much safer than fiddling with controls or trying to type in a phone number while driving.

Current voice recognition engines in in-car entertainment system work quite well when the car is not moving. But the noise in and around the car is severe and this makes it hard to recognize speech. With BeClear Speech Enhancement, voice recognition and calling in the noisy car environment are like a face to face conversation, relaxed and effortless.

Flexible building blocks

BeClear Speech Enhancement is a single algorithm that features flexible building blocks for improving the quality of captured speech.

 BeClear super de-reverberation deploys a non-traditional method of removing reverberation from the signal, extending the range of Far Field communication up to 5 meters and beyond.


BeClear is the only solution currently on the market capable of full duplex, multi-channel echo cancellation.


Fast tracking beamforming continuously focuses the device’s microphones on the person speaking for superior Dynamic Noise Suppression


Dynamic Noise suppression accurately identifies and removes background noises to leave only the sound of the person speaking for clearer conversations


Smart far-end signal enhancements improve the clarity of incoming voice calls to improve the user’s experience of your product


 

Building blocks explained

Read more

Building blocks explained

Super de-reverberation

When the person speaking is far away from the microphone, sound reflected from the walls of the room can create an annoying echo effect called reverberation.

Ideally, only sound coming directly from the person speaking should be captured. Removing the reflected sounds can therefore improve the sound quality. However, our unique solution exploits certain reflections to boost the direct sound level enabling people far from the microphone to be heard clearly without having to raise their voice.

• Reverberation from reflections is minimized even beyond 5 m
• Patented method to effectively exploit reflective sound paths and thereby improve far-field voice path
• Speech clarity index greater than 7 dB (exceeds requirements of Skype)

Full-duplex Multi-Channel Echo Cancellation
Echo cancellation is a major challenge for small form factor devices where the speakers and microphones are closely spaced in the same housing. And today’s mobile devices often have multiple speakers and microphones making the challenge even greater.

• BeClear is the  only solution on the market able to maintain full duplex multi-channel echo cancellation with high suppression
• Delivers 50-60 dB suppression for moderate non-linearities
• Scalable to devices with more than two loudspeakers

Fast tracking beamforming

Beamforming is widely used to focus the microphones on the speaker to perform noise suppression and other voice enhancements. However if the speaker moves, the beam can have a hard time tracking the speaker, in particular at larger distances, leading to loss of quality.

• Audio beam positioning between person(s) and device stays optimal

– Instant (<16 ms) adaptation of beam
– Allows for multiple beams in parallel
– Regardless of movement of device and/or person(s) speaking

• Enables consumer requirement for conferencing use case with mobile

Dynamic Noise Suppression

There are many different types of background noise that should be removed to clearly hear the person speaking.

BeClear provides simple parameters that allow the algorithm to be optimized for the best subjective performance.

• Highly accurate identification of desired and undesired sources
• 40 dB suppression for diffuse and stationary noise*
• Fast adaptation for stationary and non-stationary noise sources
• Robust even under cafeteria noise conditions; a highly challenging environment because of multiple, competing voice sources

* 12% better than competition on standardized 3QUEST test

Far-end signal enhancement

Many of the enhancements affect the quality of the captured voice signal. To ensure clear communication for the user as well as the call recipient, BeClear includes a range of Far-end signal enhancements to improve the local user experience.

• Improves voice quality of incoming signal

– Removes residual echo and noise
– Reduces reverberation components
– Optimized to prevent tandem effects

• Delivers immediate value to your customers


 

close

Hear the difference

Hear the difference for yourself by listening to the samples below with and without speech enhancement.

 

Demonstration

Read more

Demonstration

Noise Suppression

In the example below you can experience how Philips noise suppression is able to almost completely remove the background noise. In the examples the (unprocessed) microphone signal is provided as well as the processed output. Suppression is set to 30dB. Different types and levels of noise are added.

Babble Noise

Noise Level Microphone input After processing
50dB
60dB

Stationary Noise

Noise Level Microphone input After processing
50dB
60dB

De-reverberation

In the example below you can experience how Philips de-reverberation suppresses the reverberant echoes. In the example the (unprocessed) microphone signal is provided as well as the processed output. The user is at a 4 meter distance in a room with a reverberation time T60 = 800ms.

Microphone input After processing

AEC Double talk performance

In the example below you can experience how the Philips Acoustic Echo Canceller is able to maintain full duplex. In the example the (unprocessed) microphone signal is provided as well as the processed output. You will hear that the echo resulting from the far-end (female) voice is suppressed in the processed signal whereas the near-end (male) voice remains intelligible without in double talk. The users are at a 4 meter distance in a room with a reverberation time T60 = 800ms. The far end loudspeaker is at a (center-center) distance of 60cm from the microphone array. Note that AEC results are very tightly connected to the microphone and speaker configuration in the product design.

Microphone input After processing
close

Amazon Partner page

Market Focus and Applicability

Overall description and diagrams of proposed solution
BeClear SuperHandsFree (SHF) provides for high quality speech for communication or speech recognition under reverberant and noisy conditions (both diffuse and point-noise) at larger distances. The below figure provides an illustration of the SHF algorithm. By means of a fast tracking free-running beam (orange), the SHF algorithm generates one or more focused beams (blue) capturing the coherent sources.

More detail:

Read more

More detail:

  • The BeClear SHF algorithm allocates focused beams to coherent sources, for example to a desired speaker or a point noise source such as a radio or TV
  • Incoherent or diffuse noise sources such as an air conditioner or nearby traffic will be attenuated
  • A fast-tracking free running beam continuously adapts to the source with the highest coherent energy component
  • Each focused beam is able to autonomously track slow (e.g. head) movements of a coherent source in noisy conditions
  • In case a new source appears and all the focused beams are already allocated, the algorithm automatically re-allocates an existing focused beam to this new source based on information related to the recent activity and energies in the present focused beams
  • At the point a focused beam has properly converged to a source, signal content outside the focused beam will be suppressed, thereby increasing the SNR in each of the focused beams

The SHF algorithm will automatically adapt to the number of microphones and the geometry they are placed in. SHF can be combined with Multichannel Acoustic Echo Cancellation (AEC). In a typical configuration SHF uses 4 microphones and 1 or 2 AEC-reference signals.

As illustrated in the above figure, the algorithm provides for a customizable application layer for constructing the desired output signal based on the available beams. E.g.

  • Communication → Select the beam with the strongest coherent source
  • Voice recognition → Select the beam with the voice trigger
  • Multi-user → Tracking of multiple users

Differentiators of the BeClear technologies are
- Adaptivity of algorithms (agnostic to microphone geometry)

  • The algorithm always tries to do the best it can (e.g. independent on which microphones are blocked)

- Natural speech
- Flexibility in SW adaptability
- Platform (SoC) agnostic
- Highly configurable
- Multi-channel Acoustic Echo cancellation

  • No modification of loudspeaker signals needed
  • No need for explicit double talk detector
  • Limited divergence during double talk
  • Especially beneficial for Stereo AEC (and beyond)
  • Largest improvements for well-designed acoustic systems
  • Reduced complexity mode for bandwidth limited (LFE) signal

- Beamforming far-field and near-field

  • Far-field beamforming
  •  Non-traditional beamforming (de-reverberation)
  •  Multi-channel echo cancellation
  •  Diffuse noise suppression
  •  Point noise source suppression
  • Near-field beamforming
  •  Non-traditional beamforming
  •  Mono echo cancellation
  •  Diffuse noise suppression
  •  Point noise source suppression

- Noise suppression

  • Fast adaptation for stationary and non-stationary noise sources
  • Robust performance in challenging environments
  • Support for point noise source suppression

Target Market Segments and key value propositions
• Segment 1 – Communication: Natural speech. Especially under reverberated conditions, BeClear is able to maintain a high clarity index
• Segment 2 – Voice recognition: Wakeword and voice control/recognition under adverse noise conditions
• Segment 3 – Gaming: distinction of multiple users

Technology Partners
Cirrus Logic, (XMOS, MTK, …)

Future Markets, and roadmap plans to address target future markets
Robust point noise source suppression (negative SNR) for far field and near field. Automotive, robotics and wearables

Product Specifications

Overall description and block diagram of functional architecture

• Key SW and processing algorithm blocks

  • Multichannel AEC (tested up to 12 microphones, 8-farend channels; Module separately available
  • Beamforming (de-reverberation)
  • Postprocessing (Dynamic echo and noise suppression; Automatic Gain Control (AGC); Direction of Arrival (DoA))

• Interface details for audio signals and control signals



• Functional and Performance

  • Noise reduction methodology
  • Microphone Array support: mic number and array shape. At minimum 3 microphones, preferably at least 4 microphones. Microphone geometry agnostic. Both circular and linear arrays are supported
  • Echo cancellation channels. No modification of loudspeaker signals needed
  • Other features or functionality. DoA, Far-end signal processing,

• Software Tools, if applicable

  • Tuning tools and support
  • Tuning guide available
  • Support level depending on opportunity
close

Information and contact

For any other questions, please contact us

Contact us

Contact

Contact us

Fields with an asterisk (*) are required.

 
 
 
 
close