Developer Relations / Hey Reachy - Wake word detection Public
Target: Arduino UNO Q (Qualcomm QRB2210) Clone this project
Experiments EON Tuner
The EON Tuner helps you quickly run hyper-parameter sweeps that explore different pre-processing + model architectures optimized for your defined objectives. Clone this project to use the EON Tuner.

Target

Run #2

Arduino UNO Q (Qualcomm QRB2210)

100 ms

2097152 kB

8388608 kB

Filters

Status

DSP type

Model type

View

Data set

Variant

Sort

General

F1-score

Precision

Recall

97%
mfcc-conv2d-58a
PERFORMANCE
LATENCY
6 ms of 100 ms
RAM
36 kB of 2097152 kB
ROM
56 kB of 8388608 kB
DSP NN Unused
TIME-SERIES INPUT

2000 ms |
2000 ms |
Enabled

MFCC

0.05 | 0.025 | 32

ACCURACY (OVERALL)
CLASSIFICATION

0.005 | 100 | 97%

Type Filters Kernel Rate
conv2d 8 3 -
conv2d 16 3 -
conv2d 32 3 -
conv2d 64 3 -
dropout - - 0.5

2 12/18/2025, 6:45:31 PM

96%
mfcc-conv1d-ca5
PERFORMANCE
LATENCY
12 ms of 100 ms
RAM
49 kB of 2097152 kB
ROM
35 kB of 8388608 kB
DSP NN Unused
TIME-SERIES INPUT

2000 ms |
2000 ms |
Enabled

MFCC

0.02 | 0.01 | 32

ACCURACY (OVERALL)
CLASSIFICATION

0.002 | 300 | 96%

Type Filters Kernel Rate
Data augmentation
conv1d 8 3 -
conv1d 16 3 -
conv1d 32 3 -
dropout - - 0.25

0 12/18/2025, 5:52:30 PM

96%
mfcc-conv2d-326
PERFORMANCE
LATENCY
4 ms of 100 ms
RAM
33 kB of 2097152 kB
ROM
55 kB of 8388608 kB
DSP NN Unused
TIME-SERIES INPUT

1500 ms |
1500 ms |
Enabled

MFCC

0.05 | 0.05 | 32

ACCURACY (OVERALL)
CLASSIFICATION

0.005 | 100 | 96%

Type Filters Kernel Rate
conv2d 32 3 -
conv2d 64 3 -
dropout - - 0.5

0 12/18/2025, 5:38:37 PM

95%
mfcc-conv2d-b5a
PERFORMANCE
LATENCY
8 ms of 100 ms
RAM
44 kB of 2097152 kB
ROM
129 kB of 8388608 kB
DSP NN Unused
TIME-SERIES INPUT

1500 ms |
1500 ms |
Enabled

MFCC

0.05 | 0.025 | 40

ACCURACY (OVERALL)
CLASSIFICATION

0.005 | 100 | 95%

Type Filters Kernel Rate
Data augmentation
conv2d 16 3 -
conv2d 32 3 -
conv2d 64 3 -
conv2d 128 3 -
dropout - - 0.25

1 12/18/2025, 6:14:32 PM

95%
mfe-conv1d-0d9
PERFORMANCE
LATENCY
3 ms of 100 ms
RAM
23 kB of 2097152 kB
ROM
42 kB of 8388608 kB
DSP NN Unused
TIME-SERIES INPUT

1500 ms |
1500 ms |
Enabled

MFE

0.05 | 0.025 | 32

ACCURACY (OVERALL)
CLASSIFICATION

0.002 | 400 | 95%

Type Filters Kernel Rate
conv1d 16 3 -
conv1d 32 3 -
conv1d 64 3 -
dropout - - 0.25

2 12/18/2025, 6:40:38 PM

92%
mfe-conv1d-a75
PERFORMANCE
LATENCY
7 ms of 100 ms
RAM
45 kB of 2097152 kB
ROM
564 kB of 8388608 kB
DSP NN Unused
TIME-SERIES INPUT

1500 ms |
1500 ms |
Enabled

MFE

0.032 | 0.016 | 32

ACCURACY (OVERALL)
CLASSIFICATION

0.002 | 200 | 92%

Type Filters Kernel Rate
Data augmentation
conv1d 64 3 -
conv1d 128 3 -
conv1d 256 3 -
conv1d 512 3 -
dropout - - 0.5

0 12/18/2025, 5:54:28 PM

65%
mfe-conv2d-760
PERFORMANCE
LATENCY
7 ms of 100 ms
RAM
54 kB of 2097152 kB
ROM
57 kB of 8388608 kB
DSP NN Unused
TIME-SERIES INPUT

2000 ms |
2000 ms |
Enabled

MFE

0.02 | 0.02 | 32

ACCURACY (OVERALL)
CLASSIFICATION

0.005 | 100 | 65%

Type Filters Kernel Rate
Data augmentation
conv2d 8 3 -
conv2d 16 3 -
conv2d 32 3 -
conv2d 64 3 -
dropout - - 0.5

0 12/18/2025, 5:51:57 PM

49%
mfe-conv1d-73e
PERFORMANCE
LATENCY
10 ms of 100 ms
RAM
64 kB of 2097152 kB
ROM
181 kB of 8388608 kB
DSP NN Unused
TIME-SERIES INPUT

2000 ms |
2000 ms |
Enabled

MFE

0.02 | 0.01 | 32

ACCURACY (OVERALL)
CLASSIFICATION

0.005 | 300 | 49%

Type Filters Kernel Rate
Data augmentation
conv1d 64 3 -
conv1d 128 3 -
conv1d 256 3 -
dropout - - 0.5

1 12/18/2025, 6:25:10 PM

mfe-mobilenetv1-c0b
PERFORMANCE
LATENCY
100 ms
RAM
2097152 kB
ROM
8388608 kB
Unused
TIME-SERIES INPUT

1000 ms |
500 ms |
Enabled

MFE

0.02 | 0.01 | 40

ACCURACY (KERAS-TRANSFER-KWS)
TRANSFER LEARNING (KEYWORD SPOTTING)

0.01 | 30

MobileNetV1 0.1
0 | 0.1

0

mfe-mobilenetv2-01c
PERFORMANCE
LATENCY
100 ms
RAM
2097152 kB
ROM
8388608 kB
Unused
TIME-SERIES INPUT

1000 ms |
500 ms |
Enabled

MFE

0.02 | 0.01 | 40

ACCURACY (KERAS-TRANSFER-KWS)
TRANSFER LEARNING (KEYWORD SPOTTING)

0.01 | 30

MobileNetV2 0.35
0 | 0.1

0

mfe-mobilenetv2-6bc
PERFORMANCE
LATENCY
100 ms
RAM
2097152 kB
ROM
8388608 kB
Unused
TIME-SERIES INPUT

1000 ms |
500 ms |
Enabled

MFE

0.02 | 0.01 | 40

ACCURACY (KERAS-TRANSFER-KWS)
TRANSFER LEARNING (KEYWORD SPOTTING)

0.01 | 30

MobileNetV2 0.35
0 | 0.1

1

mfe-mobilenetv1-1e6
PERFORMANCE
LATENCY
100 ms
RAM
2097152 kB
ROM
8388608 kB
Unused
TIME-SERIES INPUT

1000 ms |
500 ms |
Enabled

MFE

0.02 | 0.01 | 40

ACCURACY (KERAS-TRANSFER-KWS)
TRANSFER LEARNING (KEYWORD SPOTTING)

0.01 | 30

MobileNetV1 0.1
0 | 0.1

2

Help & documentation Get help on the forum Send feedback Contact sales
Join us
Terms of Service Privacy policy