Model zoo performance

The Brainchip akida_models package offers a set of pre-built akida compatible models (e.g MobileNet, AkidaNet, ViT), pretrained weights for those models and training scripts. Please refer to the model zoo API reference for a complete list of the available models.
This page lists the performance of all models from the zoo reported for both Akida 1.0 and Akida 2.0. Please refer to:
  • Akida 1.0 models for models targetting the Akida Neuromorphic Processor IP 1.0 and the AKD1000 reference SoC,

  • Akida 2.0 models for models targetting the Akida Neuromorphic Processor IP 2.0,

  • Upgrading to Akida 2.0 tutorial to understand the architectural differences between 1.0 and 2.0 models and their respective workflows.

Note

The download links provided point towards standard Tensorflow Keras models that must be converted to Akida model using cnn2snn.convert.

Akida 1.0 models

For 1.0 models, 4-bit accuracy is provided and is always obtained through a QAT phase.

Note

The “8/4/4” quantization scheme stands for 8-bit weights in the input layer, 4-bit weights in other layers and 4-bit activations.

Note

The NPs column provides the minimal number of neural processors required for the model excecution on the Akida IP. The numbers given are the result of the map operation using the Minimal MapMode targetting AKD1000 reference SoC.

image_icon_ref Image domain

Classification

Architecture

Resolution

Dataset

#Params

Quantization

Top-1 accuracy

Example

Size (KB)

NPs

Download

AkidaNet 0.25

160

ImageNet

480K

8/4/4

42.58%

an_ex

403.3

20

an_160_25_dl

AkidaNet 0.5

160

ImageNet

1.4M

8/4/4

57.80%

an_ex

1089.1

24

an_160_50_dl

AkidaNet

160

ImageNet

4.4M

8/4/4

66.94%

an_ex

4061.1

68

an_160_dl

AkidaNet 0.25

224

ImageNet

480K

8/4/4

46.71%

an_ex

409.1

22

an_224_25_dl

AkidaNet 0.5

224

ImageNet

1.4M

8/4/4

61.30%

an_ex

1202.2

32

an_224_50_dl

AkidaNet

224

ImageNet

4.4M

8/4/4

69.65%

an_ex

6294.0

116

an_224_dl

AkidaNet 0.5 edge

160

ImageNet

4.0M

8/4/4

51.66%

ane_ex

2017.4

38

ane_160_dl

AkidaNet 0.5 edge

224

ImageNet

4.0M

8/4/4

54.03%

ane_ex

2130.5

46

ane_224_dl

AkidaNet 0.5

224

PlantVillage

1.1M

8/4/4

97.92%

an_pv_ex

1019.1

33

an_pv_dl

AkidaNet 0.25

96

Visual Wake Words

229K

8/4/4

84.77%

179.6

16

vww_dl

MobileNetV1 0.25

160

ImageNet

467K

8/4/4

36.05%

376.4

20

mb_160_25_dl

MobileNetV1 0.5

160

ImageNet

1.3M

8/4/4

54.59%

1007.0

24

mb_160_50_dl

MobileNetV1

160

ImageNet

4.2M

8/4/4

65.47%

3525.8

65

mb_160_dl

MobileNetV1 0.25

224

ImageNet

467K

8/4/4

39.73%

377.9

22

mb_224_25_dl

MobileNetV1 0.5

224

ImageNet

1.3M

8/4/4

58.50%

1065.3

32

mb_224_50_dl

MobileNetV1

224

ImageNet

4.2M

8/4/4

68.76%

5223.3

110

mb_224_dl

GXNOR

28

MNIST

1.6M

2/2/1

98.03%

412.8

3

gx_dl

Object detection

Architecture

Resolution

Dataset

#Params

Quantization

mAP

Example

Size (KB)

NPs

Download

YOLOv2

224

PASCAL-VOC 2007 - person and car classes

3.6M

8/4/4

41.51%

yl_voc_ex

3061.4

71

yl_voc_dl

YOLOv2

224

WIDER FACE

3.5M

8/4/4

77.63%

3053.1

71

yl_wf_dl

Regression

Architecture

Resolution

Dataset

#Params

Quantization

MAE

Example

Size (KB)

NPs

Download

VGG-like

32

UTKFace (age estimation)

458K

8/2/2

6.1791

reg_ex

138.6

6

reg_dl

Face recognition

Architecture

Resolution

Dataset

#Params

Quantization

Accuracy

Size (KB)

NPs

Download

AkidaNet 0.5

112×96

CASIA Webface face identification

2.3M

8/4/4

70.18%

1930.1

21

fid_dl

AkidaNet 0.5 edge

112×96

CASIA Webface face identification

23.6M

8/4/4

71.13%

6980.2

34

fide_dl

audio_icon_ref Audio domain

Keyword spotting

Architecture

Dataset

#Params

Quantization

Top-1 accuracy

Example

Size (KB)

NPs

Download

DS-CNN

Google speech command

22.7K

8/4/4

91.72%

kws_ex

23.1

5

kws_dl

pointcloud_icon_ref Point cloud

Classification

Architecture

Dataset

#Params

Quantization

Accuracy

Size (KB)

NPs

Download

PointNet++

ModelNet40 3D Point Cloud

602K

8/4/4

79.78%

490.9

12

p++_dl

Akida 2.0 models

For 2.0 models, both 8-bit PTQ and 4-bit QAT numbers are given. When not explicitely stated 8-bit PTQ accuracy is given as is (ie no further tuning/training, only quantization and calibration). The 4-bit QAT is the same as for 1.0.

Note

The digit for quantization scheme stands for both weights and activations bitwidth. Weights in the first layer are always quantized to 8-bit. When given, ‘edge’ means that the model backbone output (before classification layer) is quantized to 1-bit to allow Akida edge learning.

image_icon_ref Image domain

Classification

CNNs

Architecture

Resolution

Dataset

#Params

Quantization

Accuracy

Download

AkidaNet 0.25

160

ImageNet

483K

8

4

48.50%

41.60%

an_160_25_8_dl

an_160_25_4_dl

AkidaNet 0.5

160

ImageNet

1.4M

8

4

61.86%

57.93%

an_160_50_8_dl

an_160_50_4_dl

AkidaNet

160

ImageNet

4.4M

8

4

69.94%

67.23%

an_160_8_dl

an_160_4_dl

AkidaNet 0.25

224

ImageNet

483K

8

4

52.39%

46.06%

an_224_25_8_dl

an_224_25_4_dl

AkidaNet 0.5

224

ImageNet

1.4M

8

4

64.88%

61.47%

an_224_50_8_dl

an_224_50_4_dl

AkidaNet

224

ImageNet

4.4M

8

4

72.16%

70.11%

an_224_8_dl

an_224_4_dl

AkidaNet 0.5

224

PlantVillage

1.2M

8

4

99.61%

98.25%

an_pv8_dl

an_pv4_dl

AkidaNet 0.25

96

Visual Wake Words

227K

8

4

87.03%

85.80%

vww8_dl

vww4_dl

AkidaNet18

160

ImageNet

2.4M

8

64.77%

an18_160_dl

AkidaNet18

224

ImageNet

2.4M

8

67.32%

an18_224_dl

MobileNetV1 0.25

160

ImageNet

469K

8

4

45.72%

37.51%

mb_160_25_8_dl

mb_160_25_4_dl

MobileNetV1 0.5

160

ImageNet

1.3M

8

4

60.27%

54.81%

mb_160_50_8_dl

mb_160_50_4_dl

MobileNetV1

160

ImageNet

4.2M

8

4

69.02%

65.28%

mb_160_8_dl

mb_160_4_dl

MobileNetV1 0.25

224

ImageNet

469K

8

4

49.63%

42.08%

mb_224_25_8_dl

mb_224_25_4_dl

MobileNetV1 0.5

224

ImageNet

1.3M

8

4

63.65%

59.20%

mb_224_50_8_dl

mb_224_50_4_dl

MobileNetV1

224

ImageNet

4.2M

8

4

71.18%

68.52%

mb_224_8_dl

mb_224_4_dl

GXNOR

28

MNIST

1.6M

4

98.81%

gx2_dl

Transformers

Architecture

Resolution

Dataset

#Params

Quantization

Accuracy

Download

ViT

224

ImageNet

5.8M

8

73.79% 1

vit_dl

DeiT-dist

224

ImageNet

6.0M

8

74.34% 1

deitd_dl

1(1,2)

PTQ accuracy boosted with 1 epoch of QAT.

Object detection

Architecture

Resolution

Dataset

#Params

Quantization

mAP 50

mAP 75

mAP

Download

YOLOv2 (AkidaNet 0.5 backbone)

224

PASCAL-VOC 2007

3.6M

8

4

50.96%

47.21%

27.40%

23.21%

27.79%

24.66%

yl_voc8_dl

yl_voc4_dl

CenterNet (AkidaNet18 backbone)

384

PASCAL-VOC 2007

2.4M

8

70.32%

47.30%

43.88%

ce_voc_dl

YOLOv2 (AkidaNet 0.5 backbone)

224

WIDER FACE

3.6M

8

4

80.19%

78.60%

yl_wf8_dl

yl_wf4_dl

Regression

Architecture

Resolution

Dataset

#Params

Quantization

MAE

Download

VGG-like

32

UTKFace (age estimation)

458K

8

4

6.0304

5.8858

reg8_dl

reg4_dl

Face recognition

Architecture

Resolution

Dataset

#Params

Quantization

Accuracy

Download

AkidaNet 0.5

112×96

CASIA Webface face identification

2.3M

8

4

72.83%

69.79%

fid8_dl

fid4_dl

Segmentation

Architecture

Resolution

Dataset

#Params

Quantization

Binary IOU

Download

AkidaUNet 0.5

128

Portrait128

1.1M

8

0.9057 2

unet_dl

2

PTQ accuracy boosted with 1 epoch QAT.

audio_icon_ref Audio domain

Keyword spotting

Architecture

Dataset

#Params

Quantization

Top-1 accuracy

Download

DS-CNN

Google speech command

23.8K

8

4

4 + edge

92.87%

92.67%

90.61%

kws8_dl

kws4_dl

kws4e_dl

Classification

Architecture

Resolution

Dataset

#Params

Quantization

Accuracy

Download

ViT (1 block)

224

Urbansound8k

539.9K

8

97.73% 3

audio_vit_dl

3

PTQ accuracy boosted with 5 epochs QAT.

pointcloud_icon_ref Point cloud

Classification

Architecture

Dataset

#Params

Quantization

Accuracy

Download

PointNet++

ModelNet40 3D Point Cloud

605K

8

4

80.88% 4

81.56%

p++8_dl

p++4_dl

4

PTQ accuracy boosted with 5 epochs QAT.