Gerrit / plant-disease 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

94%
rgb-mobilenetv2-70a
PERFORMANCE
LATENCY
9539 ms of 100 ms
Exceeds target by 9439 ms
RAM
738 kB of 1536 kB
ROM
661 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

160 |
160

IMAGE

RGB

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

0.0005 | 20 | 94%

MobileNetV2 0.35
64 | 0.1

7/4/2022, 2:36:13 PM

88%
rgb-mobilenetv1-d6c
PERFORMANCE
LATENCY
1026 ms of 100 ms
Exceeds target by 926 ms
RAM
264 kB of 1536 kB
ROM
325 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

160 |
160

IMAGE

RGB

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

0.0005 | 20 | 88%

MobileNetV1 0.25
64 | 0.1 |

7/4/2022, 9:06:11 PM

82%
rgb-mobilenetv2-fb5
PERFORMANCE
LATENCY
6211 ms of 100 ms
Exceeds target by 6111 ms
RAM
674 kB of 1536 kB
ROM
177 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

160 |
160

IMAGE

RGB

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

0.0005 | 20 | 82%

MobileNetV2 0.05
16 | 0.1

7/4/2022, 9:23:17 PM

82%
rgb-mobilenetv2-789
PERFORMANCE
LATENCY
1144 ms of 100 ms
Exceeds target by 1044 ms
RAM
674 kB of 1536 kB
ROM
248 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

160 |
160

IMAGE

RGB

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

0.0005 | 20 | 82%

MobileNetV2 0.05
64 | 0.5 |

7/4/2022, 3:40:52 PM

76%
rgb-conv2d-84b
PERFORMANCE
LATENCY
4195 ms of 100 ms
Exceeds target by 4095 ms
RAM
52 kB of 1536 kB
ROM
60 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

64 |
64

IMAGE

RGB

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 10 | 76%

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

7/4/2022, 12:51:40 PM

51%
grayscale-conv2d-d8f
PERFORMANCE
LATENCY
7404 ms of 100 ms
Exceeds target by 7304 ms
RAM
50 kB of 1536 kB
ROM
69 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

64 |
64

IMAGE

Grayscale

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 10 | 51%

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

7/4/2022, 12:51:06 PM

21%
grayscale-mobilenetv1-b48
PERFORMANCE
LATENCY
1202 ms of 100 ms
Exceeds target by 1102 ms
RAM
264 kB of 1536 kB
ROM
312 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

160 |
160

IMAGE

Grayscale

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

0.0005 | 20 | 21%

MobileNetV1 0.25
16 | 0.5

7/4/2022, 3:30:08 PM

20%
rgb-mobilenetv1-249
PERFORMANCE
LATENCY
1013 ms of 100 ms
Exceeds target by 913 ms
RAM
207 kB of 1536 kB
ROM
229 kB of 8096 kB
DSP NN Unused
IMAGE INPUT

160 |
160

IMAGE

RGB

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

0.0005 | 20 | 20%

MobileNetV1 0.2
16 | 0.5 |

7/4/2022, 1:40:22 PM

rgb-mobilenetv2-cc7
PERFORMANCE
LATENCY
100 ms
RAM
1536 kB
ROM
8096 kB
Unused
IMAGE INPUT

160 |
160

IMAGE

RGB

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

0.0005 | 20

MobileNetV2 0.35
16 | 0.1 |

rgb-mobilenetv2-3d0
PERFORMANCE
LATENCY
100 ms
RAM
1536 kB
ROM
8096 kB
Unused
IMAGE INPUT

160 |
160

IMAGE

RGB

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

0.0005 | 20

MobileNetV2 160x160 0.35
16 | 0.5

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