Edge Impulse Experts / Continuous Gait Monitor (anomaly detection) Public
The EON Tuner helps you find the most optimal architecture for your embedded machine-learning application. Clone this project to use the EON Tuner.

Target

Try increase accuracy

Nordic nRF5340 DK (Cortex-M33 128MHz)

100 ms

512 kB

1024 kB

Filters

Status

DSP type

Model type

View

Data set

Variant

Sort

General

F1-score

Precision

Recall

100%
spectr-dense-90d
PERFORMANCE
LATENCY
3 ms of 100 ms
RAM
4 kB of 512 kB
ROM
19 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

Parameter Value
wavelet haar
wavelet-level 1

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 100%

Type Filters Kernel Rate
dense 40 - -
dense 20 - -
dropout - - 0.25

3/19/2024, 3:07:37 PM

100%
spectr-dense-914
PERFORMANCE
LATENCY
3 ms of 100 ms
RAM
4 kB of 512 kB
ROM
21 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

64

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 100%

Type Filters Kernel Rate
dense 40 - -
dense 20 - -
dense 10 - -
dropout - - 0.25

3/19/2024, 3:06:35 PM

100%
spectr-dense-9c6
PERFORMANCE
LATENCY
9 ms of 100 ms
RAM
3 kB of 512 kB
ROM
17 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

16

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 100%

Type Filters Kernel Rate
dense 40 - -
dense 20 - -
dropout - - 0.5

3/19/2024, 3:00:59 PM

100%
spectr-dense-255
PERFORMANCE
LATENCY
9 ms of 100 ms
RAM
3 kB of 512 kB
ROM
17 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

16

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 100%

Type Filters Kernel Rate
dense 40 - -
dense 20 - -
dropout - - 0.25

3/19/2024, 3:00:39 PM

100%
spectr-dense-33c
PERFORMANCE
LATENCY
4 ms of 100 ms
RAM
4 kB of 512 kB
ROM
20 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

64

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 100%

Type Filters Kernel Rate
dense 40 - -
dense 20 - -
dropout - - 0.25

3/19/2024, 2:59:14 PM

100%
spectr-dense-007
PERFORMANCE
LATENCY
3 ms of 100 ms
RAM
4 kB of 512 kB
ROM
20 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

64

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 100%

Type Filters Kernel Rate
dense 40 - -
dense 20 - -
dropout - - 0.5

3/19/2024, 2:59:08 PM

100%
spectr-dense-15f
PERFORMANCE
LATENCY
3 ms of 100 ms
RAM
4 kB of 512 kB
ROM
19 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

Parameter Value
wavelet bior1.3
wavelet-level 1

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 100%

Type Filters Kernel Rate
dense 40 - -
dense 20 - -
dropout - - 0.25

3/19/2024, 2:59:11 PM

97%
spectr-dense-d64
PERFORMANCE
LATENCY
3 ms of 100 ms
RAM
4 kB of 512 kB
ROM
19 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

Parameter Value
wavelet haar
wavelet-level 1

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 97%

Type Filters Kernel Rate
dense 40 - -
dense 20 - -
dropout - - 0.5

3/19/2024, 3:04:40 PM

97%
spectr-dense-32f
PERFORMANCE
LATENCY
9 ms of 100 ms
RAM
4 kB of 512 kB
ROM
18 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

16

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 97%

Type Filters Kernel Rate
dense 40 - -
dense 20 - -
dense 10 - -
dropout - - 0.25

3/19/2024, 3:04:26 PM

97%
spectr-dense-584
PERFORMANCE
LATENCY
9 ms of 100 ms
RAM
4 kB of 512 kB
ROM
18 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

16

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 97%

Type Filters Kernel Rate
dense 40 - -
dense 20 - -
dense 10 - -
dropout - - 0.5

3/19/2024, 3:03:55 PM

93%
spectr-dense-0a0
PERFORMANCE
LATENCY
3 ms of 100 ms
RAM
4 kB of 512 kB
ROM
17 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

64

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 93%

Type Filters Kernel Rate
dense 20 - -
dense 10 - -
dropout - - 0.25

3/19/2024, 3:05:44 PM

72%
spectr-dense-44a
PERFORMANCE
LATENCY
3 ms of 100 ms
RAM
4 kB of 512 kB
ROM
17 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

Parameter Value
wavelet haar
wavelet-level 1

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 72%

Type Filters Kernel Rate
dense 20 - -
dense 10 - -
dropout - - 0.25

3/19/2024, 3:01:13 PM

69%
spectr-dense-172
PERFORMANCE
LATENCY
3 ms of 100 ms
RAM
4 kB of 512 kB
ROM
20 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

Parameter Value
wavelet bior1.3
wavelet-level 1

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 69%

Type Filters Kernel Rate
dense 40 - -
dense 20 - -
dense 10 - -
dropout - - 0.25

3/19/2024, 3:02:21 PM

69%
raw-dense-8ab
PERFORMANCE
LATENCY
1 ms of 100 ms
RAM
3 kB of 512 kB
ROM
28 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

RAW

1

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 69%

Type Filters Kernel Rate
dense 40 - -
dense 20 - -
dropout - - 0.25

3/19/2024, 3:09:05 PM

62%
spectr-dense-cde
PERFORMANCE
LATENCY
4 ms of 100 ms
RAM
4 kB of 512 kB
ROM
18 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

5000 ms |
5000 ms |
Enabled

SPECTRAL-ANALYSIS

64

ACCURACY (KERAS)
CLASSIFICATION

0.0005 | 30 | 62%

Type Filters Kernel Rate
dense 20 - -
dense 10 - -
dense 5 - -
dropout - - 0.25

3/19/2024, 3:02:43 PM