Audio and Speech Modules Development

AMI Technologies closely work with the customers to accelerate their Algorithm development needs. We offer the more than 400 audio and speech algorithms Off-the-shelf and rapidly develop the new algorithms in combination with these or form the scratch

Key Offerings

1. Design and develop the application specific algorithms for example Wind Noise Reduction, Convolution Reverb etc
2. Development starts by tweaking the algorithm in the Matlab and unit test using the file I/O or specific input data
3. Porting the Matlab code to C and optimize w.r.t Memory and Cycles
4. Test the algorithm in real time on Native PC and make it to meet the given specs
5. Port that to given Evaluation kit of DSP/DSC or SoC etc
6. Create a real time demo on Native CP or evaluation kit using the AMI Designer to further evaluate by the customers
7. Optimize to fit in the available Memory and MIPS

How It Works

Our "Off-the-Shelf" modules are categorized into grouped like EQs , Cross Overs, Virtualizer, Bass Boost, Speed EQs, Auto Duckers, Filters, Scalars, Mixers, Delays, Limiters, etc. These low-level modules can be combined to provide increased complexity layouts using GUI Designer and generate API for it. Most Audio modules are designed to operate on mono or interleaved stereo audio channels, which allowing mono, and stereo signals to be supported in a natural manner. Some modules incorporate built-in smoothing on a sample-by-sample or block-by-block basis yielding "click less" operation. It supports collection of optimized real-time Audio processing functions sufficient for developing innovative Audio products.

Repository of Modules