Firmware Devs @fw_devs

Firmware-Developments is managed by veterans in audio and speech signal processing. We help you to design and optimise the performance of your system. firmware-developments.com ANTIBES, FRANCE Joined March 2013

Tweets

197
Followers

100
Following

203
Likes

20

Firmware Devs @fw_devs

3 years ago

Currently tuning a pitch detector for a fast TSM algorithm (Time-Scale Modification of speech and audio signal).

0 0 0 0 0

Edge Impulse @EdgeImpulse

4 years ago

This new @ArmSoftwareDev demo illustrates how to deploy a tinyML keyword spotting application with echo cancellation on an @Arm Cortex-M55 processor: bit.ly/3pk8OTT

0 5 7 0 0

Download Image

Arm Software Developers @ArmSoftwareDev

4 years ago

👩‍💻 New Demo 👨‍💻 Deploy a #tinyML keyword spotting application with echo cancellation on the @Arm Cortex-M55 processor: bit.ly/3sf1MBB

0 3 2 0 0

Download Gif

Firmware Devs @fw_devs

4 years ago

C is the greenest programming language 😎 hackaday.com/2021/11/18/c-i…

0 0 2 0 0

Firmware Devs @fw_devs

4 years ago

Demonstration of our low-complexity dynamic range compressor : firmware-developments.com/WP/low-complex…

0 0 1 0 0

Firmware Devs @fw_devs

5 years ago

We are registered in “la place business" organized in Nice for entrepreneurs and startups laplacebusiness.com/entreprise/fir…

0 0 0 0 0

Firmware Devs @fw_devs

5 years ago

Linguistics : quick reading book (170 pages with many illustrations) on languages, with a good historical perspective on theories

0 0 0 0 0

Download Image

Firmware Devs @fw_devs

5 years ago

Current sources of inspiration (sometimes we stop coding 😉)

0 0 1 0 0

Download Image

Firmware Devs @fw_devs

5 years ago

Currently developing "smart features" and auto-calibration schemes to convert raw accelerometer data from large files : two months of continuous recording of several elevators. This must fit in a Cortex-M4, so deep-learning is not an option.

0 0 1 0 0

Laurent Le Faucheur @armllf

6 years ago

Cortex-M55 A massive performance boost for DSP/ML: arm.com/company/news/2…

0 1 1 0 0

Firmware Devs @fw_devs

6 years ago

Antibes Sun, Sand, Snow And proud to work here

0 1 2 0 0

Download Image

Firmware Devs @fw_devs

6 years ago

Context understanding is a must for speech recognition. Alexander Waibel (CMU) on LVCSR: “So you have things like ‘euthanasia’, which could be ‘youth in Asia’. Or if you say ‘Give me a new display’ it could be understood as ‘give me a nudist play’.”

0 0 0 0 0

Firmware Devs @fw_devs

6 years ago

We just ported our web-pages under Wordpress. More content will come in the area of speech detection and preprocessing, after our annual general meeting. firmware-developments.com

0 0 0 0 0

Firmware Devs @fw_devs

6 years ago

Currently benchmarking our formants and pitch estimator on the TIMIT database.

0 0 3 0 0

Download Image

Firmware Devs @fw_devs

6 years ago

A low-power design trick for always-on sound triggers Activate the PDM clock by bursts of 10ms (MIC sensitivity is recovered after 8ms), analyse the energy of the last 2ms and decide to maintain the PDM clock or go in idle.

0 4 7 0 1

Firmware Devs @fw_devs

6 years ago

Currently tuning voice biometrics algorithms, which will avoid you repeating the wake-up keyword, once your smart-speaker device has already recognised you. This will let you speak like in a normal conversation.

0 0 0 0 0

Firmware Devs @fw_devs

6 years ago

Advanced voice services with TTS for theaters, were commercially available in 1989 around Lannion, thanks to OrangeLabs/CNET-LAA researchers 😉 Amazon sued by early voice tech developers who allege that company infringed on patents geekwire.com/2019/amazon-su… via @geekwire

0 0 0 0 0

Firmware Devs @fw_devs

6 years ago

Many voice services were already deployed in Britanny in years of 1989. People seems to have "rediscovered" the features 20 years later... Check out VB Assets vs. Amazon on Scribd: @Scribd ReadMore scribd.com/document/42015…