• 中文
  • 3.8.1

    • J6P/H 模型性能Benchmark

    • 测试开发板：J6P

    • 模型来源：OE包内 samples/ucp_tutorial/dnn/ai_benchmark/j6 路径下的模型

    • 运行环境：Linux

    模型精度

    MODEL NAMEINPUT SIZEACCURACYDataset
    ResNet50

    1x3x224x224

    Top1:

    0.7703(FLOAT)/0.7661(INT8)

    		ImageNet
    GoogleNet

    1x3x224x224

    Top1:

    0.7018(FLOAT)/0.6995(INT8)

    		ImageNet
    EfficientNet_Lite1

    1x240x240x3

    Top1:

    0.7652(FLOAT)/0.7602(INT8)

    		ImageNet
    EfficientNet_Lite2

    1x260x260x3

    Top1:

    0.7734(FLOAT)/0.7696(INT8)

    		ImageNet
    EfficientNet_Lite3

    1x280x280x3

    Top1:

    0.7917(FLOAT)/0.7885(INT8)

    		ImageNet
    EfficientNet_Lite4

    1x300x300x3

    Top1:

    0.8063(FLOAT)/0.8041(INT8)

    		ImageNet
    Vargconvnet

    1x3x224x224

    Top1:

    0.7793(FLOAT)/0.7765(INT8)

    		ImageNet
    Efficientnasnet_m

    1x3x300x300

    Top1:

    0.7935(FLOAT)/0.7923(INT8)

    		ImageNet
    Efficientnasnet_s

    1x3x280x280

    Top1:

    0.7441(FLOAT)/0.7516(INT8)

    		ImageNet
    ResNet18

    1x3x224x224

    Top1:

    0.6976(FLOAT)/0.6948(INT8)

    		ImageNet
    YOLOv2_Darknet19

    1x3x608x608

    [IoU=0.50:0.95]=

    0.2760(FLOAT)/0.2707(INT8)

    		COCO
    YOLOv3_Darknet53

    1x3x416x416

    [IoU=0.50:0.95]=

    0.3370(FLOAT)/0.3370(INT8)

    		COCO
    YOLOv5x_v2.0

    1x3x672x672

    [IoU=0.50:0.95]=

    0.4810(FLOAT)/0.4670(INT8)

    		COCO
    Centernet_resnet101

    1x3x512x512

    [IoU=0.50:0.95]=

    0.3420(FLOAT)/0.3270(INT8)

    		COCO
    YOLOv3_VargDarknet

    1x3x416x416

    [IoU=0.50:0.95]=

    0.3280(FLOAT)/0.3260(INT8)

    		COCO
    Deeplabv3plus_efficientnetb0

    1x3x1024x2048

    mIoU:

    0.7630(FLOAT)/0.7569(INT8)

    		Cityscapes
    Fastscnn_efficientnetb0

    1x3x1024x2048

    mIoU:

    0.6997(FLOAT)/0.6909(INT8)

    		Cityscapes
    Deeplabv3plus_efficientnetm1

    1x3x1024x2048

    mIoU:

    0.7794(FLOAT)/0.7756(INT8)

    		Cityscapes
    Deeplabv3plus_efficientnetm2

    1x3x1024x2048

    mIoU:

    0.7882(FLOAT)/0.7854(INT8)

    		Cityscapes
    Bev_lss_efficientnetb0_multitask

    image:

    6x3x256x704

    points(0&1):

    10x128x128x2

    NDS:

    0.3005(FLOAT)/0.3010(INT8)

    MeanIOU:

    0.5180(FLOAT)/0.5172(INT8)

    mAP:

    0.2062(FLOAT)/0.2047(INT8)

    		Nuscenes
    Detr3d_efficientnetb3

    coords(0-3):

    6x4x256x2

    image:

    6x3x512x1408

    masks:

    1x4x256x24

    NDS:

    0.3304(FLOAT)/0.3306(INT8)

    mAP:

    0.2753(FLOAT)/0.2742(INT8)

    		Nuscenes
    Petr_efficientnetb3

    image:

    6x3x512x1408

    pos_embed:

    1x96x44x256

    NDS:

    0.3765(FLOAT)/0.3748(INT8)

    mAP:

    0.3038(FLOAT)/0.2942(INT8)

    		Nuscenes
    Bevformer_tiny_resnet50_detection

    img:

    6x3x480x800

    prev_bev:

    1x2500x256

    prev_bev_ref:

    1x50x50x2

    queries_rebatch_grid:

    6x20x32x2

    restore_bev_grid:

    1x100x50x2

    reference_points_rebatch:

    6x640x4x2

    bev_pillar_counts:

    1x2500x1

    NDS:

    0.3713(FLOAT)/0.3701(INT8)

    mAP:

    0.2673(FLOAT)/0.2641(INT8)

    		Nuscenes
    Flashocc_henet_lss_occ3d_nuscenes

    img:

    6x3x512x960

    points:

    10x128x128x2

    points_depth:

    10x128x128x2

    mIoU:

    0.3675(FLOAT)/0.3685(INT8)

    		Nuscenes
    Centerpoint_pointpillar

    points:

    300000x5

    voxel_feature:

    1x5x20x40000

    coors:

    40000x4

    NDS:

    0.5832(FLOAT)/0.5820(INT8)

    mAP:

    0.4804(FLOAT)/0.4783(INT8)

    		Nuscenes
    FCOS3D_efficientnetb0

    1x3x512x896

    NDS:

    0.3061(FLOAT)/0.3022(INT8)

    mAP:

    0.2133(FLOAT)/0.2098(INT8)

    		nuscenes
    Ganet_mixvargenet

    1x3x320x800

    F1Score:

    0.7949(FLOAT)/0.7884(INT8)

    		CuLane
    Deformable_detr_resnet50

    1x3x800x1333

    [IoU=0.50:0.95]=

    0.4413(FLOAT)/0.4499(INT8)

    		MS COCO
    Unet_mobilenetv1

    1x3x1024x2048

    mIoU:

    0.6802(FLOAT)/0.6764(INT8)

    		Cityscapes
    Maptroe_henet_tinym_bevformer

    img:

    6x3x480x800

    osm_mask:

    1x1x50x100

    queries_rebatch_grid:

    6x20x100x2

    restore_bev_grid:

    1x100x100x2

    reference_points_rebatch:

    6x2000x4x2

    bev_pillar_counts:

    1x5000x1

    mAP:

    0.6632(FLOAT)/0.6566(INT8)

    		Nuscenes
    Qcnet_oe

    valid_mask:

    1x30x10

    valid_mask_a2a:

    1x10x30x30

    agent_type:

    1x30x1

    x_a_cur:

    1x1x30x1,1x1x30x1,1x1x30x1,1x1x30x1

    r_pl2a_cur:

    1x1x30x80,1x1x30x80,1x1x30x80

    r_t_cur:

    1x1x30x6,1x1x30x6,1x1x30x6,1x1x30x6

    r_a2a_cur:

    1x1x30x30,1x1x30x30,1x1x30x30

    x_a_mid_emb:

    1x30x2x128

    x_a:

    1x30x6x128

    pl_type,is_intersection:

    1x80

    r_pl2pl:

    1x1x80x80,1x1x80x80,1x1x80x80

    r_pt2pl:

    1x1x80x50,1x1x80x50,1x1x80x50

    mask_pl2pl:

    1x80x80

    magnitude,pt_type,side,mask:

    1x80x50

    mask_a2m:

    1x30x30

    mask_dst:

    1x30x1

    type_pl2pl:

    1x80x80

    hitrate:

    0.8026(FLOAT)/0.7979(INT8)

    		Argoverse 2

    模型性能

    注解

    在进行性能评测前，请参考如下方式，通过环境变量设置模型推理中CPU算子推理线程池的线程数量为12：

    export HB_UCP_ENABLE_CPU_BACKEND_CORE_NUM=12
    MODEL NAMEINPUT SIZELatency(ms)FPSFPS Configuration
    ResNet50

    1x3x224x224

    		0.5826553.92thread_num:10
    GoogleNet

    1x3x224x224

    		0.38118761.00thread_num:12
    EfficientNet_Lite1

    1x240x240x3

    		0.38418139.78thread_num:12
    EfficientNet_Lite2

    1x260x260x3

    		0.44714117.23thread_num:10
    EfficientNet_Lite3

    1x280x280x3

    		0.51311575.29thread_num:10
    EfficientNet_Lite4

    1x300x300x3

    		0.6538334.94thread_num:10
    Vargconvnet

    1x3x224x224

    		0.52112421.03thread_num:12
    Efficientnasnet_m

    1x3x300x300

    		0.5819270.39thread_num:10
    Efficientnasnet_s

    1x3x280x280

    		0.37619655.71thread_num:12
    ResNet18

    1x3x224x224

    		0.38914388.70thread_num:12
    YOLOv2_Darknet19

    1x3x608x608

    		2.2241988.29thread_num:8
    YOLOv3_Darknet53

    1x3x416x416

    		2.4411765.76thread_num:8
    YOLOv5x_v2.0

    1x3x672x672

    		8.190495.13thread_num:8
    Centernet_resnet101

    1x3x512x512

    		2.4391793.76thread_num:8
    YOLOv3_VargDarknet

    1x3x416x416

    		1.6252696.32thread_num:8
    Deeplabv3plus_efficientnetb0

    1x3x1024x2048

    		2.9541468.90thread_num:8
    Fastscnn_efficientnetb0

    1x3x1024x2048

    		1.7392647.01thread_num:8
    Deeplabv3plus_efficientnetm1

    1x3x1024x2048

    		5.123817.97thread_num:8
    Deeplabv3plus_efficientnetm2

    1x3x1024x2048

    		7.326562.40thread_num:8
    Bev_lss_efficientnetb0_multitask

    image:

    6x3x256x704

    points(0&1):

    10x128x128x2

    		3.0561452.42thread_num:8
    Detr3d_efficientnetb3

    coords(0-3):

    6x4x256x2

    image:

    6x3x512x1408

    masks:

    1x4x256x24

    		15.104257.70thread_num:8
    Petr_efficientnetb3

    image:

    6x3x512x1408

    pos_embed:

    1x96x44x256

    		21.426188.17thread_num:8
    Bevformer_tiny_resnet50_detection

    img:

    6x3x480x800

    prev_bev:

    1x2500x256

    prev_bev_ref:

    1x50x50x2

    queries_rebatch_grid:

    6x20x32x2

    restore_bev_grid:

    1x100x50x2

    reference_points_rebatch:

    6x640x4x2

    bev_pillar_counts:

    1x2500x1

    		14.104276.96thread_num:8
    Flashocc_henet_lss_occ3d_nuscenes

    img:

    6x3x512x960

    points:

    10x128x128x2

    points_depth:

    10x128x128x2

    		5.506744.62thread_num:8
    Centerpoint_pointpillar

    points:

    300000x5

    voxel_feature:

    1x5x20x40000

    coors:

    40000x4

    		7.4811079.91thread_num:16
    FCOS3D_efficientnetb0

    1x3x512x896

    		1.4663593.00thread_num:10
    Ganet_mixvargenet

    1x3x320x800

    		0.55112404.70thread_num:12
    Deformable_detr_resnet50

    1x3x800x1333

    		77.00129.32thread_num:8
    Unet_mobilenetv1

    1x3x1024x2048

    		0.8846399.42thread_num:10
    Maptroe_henet_tinym_bevformer

    img:

    6x3x480x800

    osm_mask:

    1x1x50x100

    queries_rebatch_grid:

    6x20x100x2

    restore_bev_grid:

    1x100x100x2

    reference_points_rebatch:

    6x2000x4x2

    bev_pillar_counts:

    1x5000x1

    		6.779583.17thread_num:8
    Qcnet_oe

    valid_mask:

    1x30x10

    valid_mask_a2a:

    1x10x30x30

    agent_type:

    1x30x1

    x_a_cur:

    1x1x30x1,1x1x30x1,1x1x30x1,1x1x30x1

    r_pl2a_cur:

    1x1x30x80,1x1x30x80,1x1x30x80

    r_t_cur:

    1x1x30x6,1x1x30x6,1x1x30x6,1x1x30x6

    r_a2a_cur:

    1x1x30x30,1x1x30x30,1x1x30x30

    x_a_mid_emb:

    1x30x2x128

    x_a:

    1x30x6x128

    pl_type,is_intersection:

    1x80

    r_pl2pl:

    1x1x80x80,1x1x80x80,1x1x80x80

    r_pt2pl:

    1x1x80x50,1x1x80x50,1x1x80x50

    mask_pl2pl:

    1x80x80

    magnitude,pt_type,side,mask:

    1x80x50

    mask_a2m:

    1x30x30

    mask_dst:

    1x30x1

    type_pl2pl:

    1x80x80

    		2.6021653.80thread_num:8