Brainchip / Image Classification - Deck of Cards - BrainChip Akida - Edge Learning 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

No name set

BrainChip AKD1000 MINI PCIe Board

100 ms

524288 kB

16384 kB

Filters

Status

DSP type

Model type

View

Data set

Variant

Sort

General

F1-score

Precision

Recall

94%
rgb-conv2d-9c4
PERFORMANCE
LATENCY
14131 ms of 100 ms
Exceeds target by 14031 ms
RAM
3927 kB of 524288 kB
ROM
20814 kB of 16384 kB
Exceeds target by 4430 kB
DSP NN Unused
IMAGE INPUT

224 |
224

IMAGE

RGB

CLASSIFICATION

0.0005 | 10 | 94%

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

1/27/2023, 7:56:02 PM

93%
rgb-conv2d-d43
PERFORMANCE
LATENCY
16699 ms of 100 ms
Exceeds target by 16599 ms
RAM
2009 kB of 524288 kB
ROM
3050 kB of 16384 kB
DSP NN Unused
IMAGE INPUT

160 |
160

IMAGE

RGB

CLASSIFICATION

0.0005 | 10 | 93%

Type Filters Kernel Rate
conv2d 16 3 -
conv2d 32 3 -
conv2d 64 3 -
conv2d 128 3 -
dropout - - 0.5

1/27/2023, 7:53:52 PM

93%
rgb-conv2d-774
PERFORMANCE
LATENCY
14131 ms of 100 ms
Exceeds target by 14031 ms
RAM
3927 kB of 524288 kB
ROM
20814 kB of 16384 kB
Exceeds target by 4430 kB
DSP NN Unused
IMAGE INPUT

224 |
224

IMAGE

RGB

CLASSIFICATION

0.0005 | 10 | 93%

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

1/27/2023, 7:56:08 PM

92%
rgb-conv2d-36a
PERFORMANCE
LATENCY
15643 ms of 100 ms
Exceeds target by 15543 ms
RAM
1449 kB of 524288 kB
ROM
4199 kB of 16384 kB
DSP NN Unused
IMAGE INPUT

96 |
96

IMAGE

RGB

CLASSIFICATION

0.0005 | 10 | 92%

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

1/27/2023, 7:49:51 PM

92%
grayscale-conv2d-732
PERFORMANCE
LATENCY
20910 ms of 100 ms
Exceeds target by 20810 ms
RAM
3928 kB of 524288 kB
ROM
10498 kB of 16384 kB
DSP NN Unused
IMAGE INPUT

224 |
224

IMAGE

Grayscale

CLASSIFICATION

0.0005 | 10 | 92%

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

1/27/2023, 7:55:22 PM

91%
grayscale-conv2d-b1c
PERFORMANCE
LATENCY
11723 ms of 100 ms
Exceeds target by 11623 ms
RAM
3927 kB of 524288 kB
ROM
20813 kB of 16384 kB
Exceeds target by 4429 kB
DSP NN Unused
IMAGE INPUT

224 |
224

IMAGE

Grayscale

CLASSIFICATION

0.0005 | 10 | 91%

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

1/27/2023, 7:55:01 PM

90%
grayscale-conv2d-b48
PERFORMANCE
LATENCY
14758 ms of 100 ms
Exceeds target by 14658 ms
RAM
1449 kB of 524288 kB
ROM
4197 kB of 16384 kB
DSP NN Unused
IMAGE INPUT

96 |
96

IMAGE

Grayscale

CLASSIFICATION

0.0005 | 10 | 90%

Type Filters Kernel Rate
conv2d 32 3 -
conv2d 64 3 -
conv2d 128 3 -
dropout - - 0.5

1/27/2023, 7:53:12 PM

90%
grayscale-conv2d-ecc
PERFORMANCE
LATENCY
13946 ms of 100 ms
Exceeds target by 13846 ms
RAM
2567 kB of 524288 kB
ROM
13660 kB of 16384 kB
DSP NN Unused
IMAGE INPUT

128 |
128

IMAGE

Grayscale

CLASSIFICATION

0.0005 | 10 | 90%

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

1/27/2023, 7:50:25 PM

85%
rgb-conv2d-352
PERFORMANCE
LATENCY
22637 ms of 100 ms
Exceeds target by 22537 ms
RAM
1451 kB of 524288 kB
ROM
3444 kB of 16384 kB
DSP NN Unused
IMAGE INPUT

96 |
96

IMAGE

RGB

CLASSIFICATION

0.0005 | 10 | 85%

Type Filters Kernel Rate
conv2d 32 3 -
conv2d 64 3 -
conv2d 128 3 -
conv2d 256 3 -
dropout - - 0.5

1/27/2023, 7:55:44 PM

84%
grayscale-conv2d-29c
PERFORMANCE
LATENCY
21753 ms of 100 ms
Exceeds target by 21653 ms
RAM
1451 kB of 524288 kB
ROM
3442 kB of 16384 kB
DSP NN Unused
IMAGE INPUT

96 |
96

IMAGE

Grayscale

CLASSIFICATION

0.0005 | 10 | 84%

Type Filters Kernel Rate
conv2d 32 3 -
conv2d 64 3 -
conv2d 128 3 -
conv2d 256 3 -
dropout - - 0.5

1/27/2023, 7:48:00 PM