Stanton / 10n2 Public
Target: Sony Spresense (Cortex-M4F 156MHz) 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

No name set

Sony Spresense (Cortex-M4F 156MHz)

100 ms

1536 kB

8096 kB

Filters

Status

DSP type

Model type

View

Data set

Variant

Sort

General

F1-score

Precision

Recall

100%
grayscale-mobilenetv2-260
PERFORMANCE
LATENCY
449 ms of 100 ms
Exceeds target by 349 ms
RAM
297 kB of 1536 kB
ROM
223 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

96 |
96

IMAGE

Grayscale

ACCURACY (KERAS-TRANSFER-IMAGE)
TRANSFER LEARNING (IMAGES)

0.0005 | 20 | 100%

MobileNetV2 0.1
16 | 0.1 |

7/19/2022, 12:59:47 PM

99%
grayscale-mobilenetv1-f8d
PERFORMANCE
LATENCY
830 ms of 100 ms
Exceeds target by 730 ms
RAM
111 kB of 1536 kB
ROM
238 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

96 |
96

IMAGE

Grayscale

ACCURACY (KERAS-TRANSFER-IMAGE)
TRANSFER LEARNING (IMAGES)

0.0005 | 20 | 99%

MobileNetV1 0.2
64 | 0.1

7/19/2022, 1:01:26 PM

99%
rgb-mobilenetv2-169
PERFORMANCE
LATENCY
2998 ms of 100 ms
Exceeds target by 2898 ms
RAM
287 kB of 1536 kB
ROM
234 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

96 |
96

IMAGE

RGB

ACCURACY (KERAS-TRANSFER-IMAGE)
TRANSFER LEARNING (IMAGES)

0.0005 | 20 | 99%

MobileNetV2 0.05
64 | 0.5

7/19/2022, 12:58:38 PM

99%
rgb-mobilenetv2-79e
PERFORMANCE
LATENCY
3085 ms of 100 ms
Exceeds target by 2985 ms
RAM
297 kB of 1536 kB
ROM
223 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

96 |
96

IMAGE

RGB

ACCURACY (KERAS-TRANSFER-IMAGE)
TRANSFER LEARNING (IMAGES)

0.0005 | 20 | 99%

MobileNetV2 0.1
16 | 0.1

7/19/2022, 1:06:26 PM

97%
grayscale-conv2d-5ea
PERFORMANCE
LATENCY
271 ms of 100 ms
Exceeds target by 171 ms
RAM
20 kB of 1536 kB
ROM
60 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

32 |
32

IMAGE

Grayscale

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 10 | 97%

Type Filters Kernel Rate
conv2d 8 3 -
conv2d 16 3 -
conv2d 32 3 -
conv2d 64 3 -
dropout - - 0.25

7/19/2022, 1:04:46 PM

95%
rgb-mobilenetv2-fac
PERFORMANCE
LATENCY
505 ms of 100 ms
Exceeds target by 405 ms
RAM
297 kB of 1536 kB
ROM
223 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

96 |
96

IMAGE

RGB

ACCURACY (KERAS-TRANSFER-IMAGE)
TRANSFER LEARNING (IMAGES)

0.0005 | 20 | 95%

MobileNetV2 0.1
16 | 0.5 |

7/19/2022, 12:55:50 PM

95%
rgb-mobilenetv2-bc4
PERFORMANCE
LATENCY
3247 ms of 100 ms
Exceeds target by 3147 ms
RAM
297 kB of 1536 kB
ROM
284 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

96 |
96

IMAGE

RGB

ACCURACY (KERAS-TRANSFER-IMAGE)
TRANSFER LEARNING (IMAGES)

0.0005 | 20 | 95%

MobileNetV2 0.1
64 | 0.5 |

7/19/2022, 1:05:25 PM

93%
rgb-mobilenetv1-da3
PERFORMANCE
LATENCY
5175 ms of 100 ms
Exceeds target by 5075 ms
RAM
111 kB of 1536 kB
ROM
238 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

96 |
96

IMAGE

RGB

ACCURACY (KERAS-TRANSFER-IMAGE)
TRANSFER LEARNING (IMAGES)

0.0005 | 20 | 93%

MobileNetV1 0.2
64 | 0.5 |

7/19/2022, 1:08:53 PM

57%
grayscale-mobilenetv1-170
PERFORMANCE
LATENCY
2472 ms of 100 ms
Exceeds target by 2372 ms
RAM
113 kB of 1536 kB
ROM
115 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

160 |
160

IMAGE

Grayscale

ACCURACY (KERAS-TRANSFER-IMAGE)
TRANSFER LEARNING (IMAGES)

0.0005 | 20 | 57%

MobileNetV1 0.1
64 | 0.5 |

7/19/2022, 12:55:06 PM

17%
rgb-mobilenetv1-92d
PERFORMANCE
LATENCY
4392 ms of 100 ms
Exceeds target by 4292 ms
RAM
130 kB of 1536 kB
ROM
110 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

160 |
160

IMAGE

RGB

ACCURACY (KERAS-TRANSFER-IMAGE)
TRANSFER LEARNING (IMAGES)

0.0005 | 20 | 17%

MobileNetV1 0.1
16 | 0.5

7/19/2022, 1:03:04 PM

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