sklearn SVM Classification - Spectral features

Edge Impulse Inc. / sklearn SVM Classification Public

Target: Cortex-M7 216MHz Clone this project

Generate features

Raw data

Show:

Raw features

...

Label

...

Parameters

Filter

Scale axes

Scale axes should be a number

Input decimation ratio

Type

Cut-off frequency

Cut-off frequency should be a number

Order

Order should be a number

Analysis

Type

FFT length

FFT length should be a number

No. of peaks

No. of peaks should be a number

Peaks threshold

Peaks threshold should be a number

Power edges

Power edges should not be empty

Take log of spectrum?

Overlap FFT frames?

Wavelet Decomposition Level

Wavelet Decomposition Level should be a number

Wavelet

Improve low frequency resolution?

DSP result

Error: XXX

Processed features

...

Raw data

Raw features

Label

Parameters

Filter

Analysis

Autotune output

DSP result

Processed features

On-device performance

Processing time

Peak RAM usage

Error while loading processing block