Edge Impulse Inc. / Tutorial - Syntiant TinyML Circular Motion Public

Target: Cortex-M7 216MHz Clone this project

Welcome to Edge Impulse, the largest community of edge AI developers!
This is a public Edge Impulse project, use the navigation bar to see all data and models in this project; or clone to retrain or deploy to any edge device.

Tutorial - Syntiant TinyML Circular Motion

About this project

Detect clockwise circular motions using your Syntiant TinyML board. The Neural Network is trained on data from both gyroscope and accelerometer, sampled at 100 Hz.

Data (37 samples)

View all

EI_raw_2ta_inference_tankptr_7.csv.2ubkh9tv

ID: 98284113

Label: circularClockwise

Sample length: 10s

EI_raw_1ta_inference_tankptr_0.csv.2ubkj18i

ID: 98284098

Label: z_openset

Sample length: 10s

EI_raw_2ta_inference_tankptr_6.csv.2ubkh9b8

ID: 98284112

Label: circularClockwise

Sample length: 8s

EI_raw_2ta_inference_tankptr_1.csv.2ubkh7ui

ID: 98284116

Label: circularClockwise

Sample length: 7s

EI_raw_2ta_inference_tankptr_0.csv.2ubkj0lc

ID: 98284102

Label: z_openset

Sample length: 10s

EI_raw_2ta_inference_tankptr_2.csv.2ubkj1n6

ID: 98284096

Label: z_openset

Sample length: 10s

EI_raw_3ta_inference_tankptr_5.csv.2ubkhcdk

ID: 98284105

Label: circularClockwise

Sample length: 9s

EI_raw_3ta_inference_tankptr_2.csv.2ubkhbh6

ID: 98284109

Label: circularClockwise

Sample length: 9s

Model accuracy (Unoptimized (float32))

Validation set

99.8%

Test set

91.7%

On-device performance (Cortex-M7 216MHz, Quantized (int8))

Latency

7 ms.

Peak RAM usage

3.3K

Flash usage

419.9K

Run this model

On any device

Clone project

Dataset summary

Data collected

5m 56s

Sensors

gyrX, gyrY, gyrZ, accX, accY, accZ @ 100Hz

Labels

circularClockwise, z_openset

Project info

Project ID	102981
Project version	2
License	BSD 3-Clause Clear
No. of views	22,248
No. of clones	65