Skip to content

Commit fc30423

Browse files
fengyuentaufengyuentau
authored
add benchmark results on Jetson Nano Orin (#224)
1 parent bee3012 commit fc30423

File tree

3 files changed

+2000
-1388
lines changed

3 files changed

+2000
-1388
lines changed

‎benchmark/README.md‎

Lines changed: 107 additions & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -813,28 +813,128 @@ mean median min input size model
813813
126.64 125.09 110.45 [1280, 720] CRNN with ['text_recognition_CRNN_EN_2022oct_int8.onnx']
814814
```
815815
816-
<!--
816+
### Jetson Nano Orin
817817
818-
### NVIDIA Jetson Orin Nano
818+
Specs: https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-orin/
819+
- CPU: 6-core Arm® Cortex®-A78AE v8.2 64-bit CPU, 1.5MB L2 + 4MB L3
820+
- GPU: 1024-core NVIDIA Ampere architecture GPU with 32 Tensor Cores, max freq 625MHz
819821
820822
CPU:
821823
822824
```
823-
59.30 58.45 57.90 [1280, 720] VitTrack with ['object_tracking_vittrack_2023sep.onnx']
825+
$ python3 benchmark.py --all
826+
Benchmarking ...
827+
backend=cv.dnn.DNN_BACKEND_OPENCV
828+
target=cv.dnn.DNN_TARGET_CPU
829+
mean median min input size model
830+
2.59 2.62 2.50 [160, 120] YuNet with ['face_detection_yunet_2023mar.onnx']
831+
2.98 2.97 2.50 [160, 120] YuNet with ['face_detection_yunet_2023mar_int8.onnx']
832+
20.05 24.76 19.75 [150, 150] SFace with ['face_recognition_sface_2021dec.onnx']
833+
31.84 32.72 19.75 [150, 150] SFace with ['face_recognition_sface_2021dec_int8.onnx']
834+
9.15 9.22 9.04 [112, 112] FacialExpressionRecog with ['facial_expression_recognition_mobilefacenet_2022july.onnx']
835+
14.33 15.35 9.04 [112, 112] FacialExpressionRecog with ['facial_expression_recognition_mobilefacenet_2022july_int8.onnx']
836+
15.00 15.17 14.80 [224, 224] MPHandPose with ['handpose_estimation_mediapipe_2023feb.onnx']
837+
18.37 18.63 14.80 [224, 224] MPHandPose with ['handpose_estimation_mediapipe_2023feb_int8.onnx']
838+
24.86 25.09 24.12 [192, 192] PPHumanSeg with ['human_segmentation_pphumanseg_2023mar.onnx']
839+
30.17 34.51 24.12 [192, 192] PPHumanSeg with ['human_segmentation_pphumanseg_2023mar_int8.onnx']
840+
18.47 18.55 18.23 [224, 224] MobileNet with ['image_classification_mobilenetv1_2022apr.onnx']
841+
17.08 17.30 15.80 [224, 224] MobileNet with ['image_classification_mobilenetv2_2022apr.onnx']
842+
21.26 15.89 15.80 [224, 224] MobileNet with ['image_classification_mobilenetv1_2022apr_int8.onnx']
843+
23.19 24.15 15.80 [224, 224] MobileNet with ['image_classification_mobilenetv2_2022apr_int8.onnx']
844+
102.30 101.90 101.44 [224, 224] PPResNet with ['image_classification_ppresnet50_2022jan.onnx']
845+
142.33 146.24 101.44 [224, 224] PPResNet with ['image_classification_ppresnet50_2022jan_int8.onnx']
846+
39.91 39.01 38.46 [320, 240] LPD_YuNet with ['license_plate_detection_lpd_yunet_2023mar.onnx']
847+
51.35 50.70 38.46 [320, 240] LPD_YuNet with ['license_plate_detection_lpd_yunet_2023mar_int8.onnx']
848+
125.31 126.50 121.92 [416, 416] NanoDet with ['object_detection_nanodet_2022nov.onnx']
849+
132.95 133.67 121.92 [416, 416] NanoDet with ['object_detection_nanodet_2022nov_int8.onnx']
850+
400.91 430.48 384.87 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
851+
476.63 509.48 384.87 [640, 640] YoloX with ['object_detection_yolox_2022nov_int8.onnx']
852+
19.16 19.91 18.04 [1280, 720] VitTrack with ['object_tracking_vittrack_2023sep.onnx']
853+
27.73 26.93 26.72 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
854+
35.16 41.14 26.72 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb_int8.onnx']
855+
33.05 33.18 32.67 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
856+
93.58 94.02 92.36 [128, 256] YoutuReID with ['person_reid_youtu_2021nov.onnx']
857+
119.80 153.20 92.36 [128, 256] YoutuReID with ['person_reid_youtu_2021nov_int8.onnx']
858+
31.51 32.19 30.69 [256, 256] MPPose with ['pose_estimation_mediapipe_2023mar.onnx']
859+
3.53 3.53 3.51 [100, 100] WeChatQRCode with ['detect_2021nov.prototxt', 'detect_2021nov.caffemodel', 'sr_2021nov.prototxt', 'sr_2021nov.caffemodel']
860+
78.10 77.77 77.17 [640, 480] PPOCRDet with ['text_detection_cn_ppocrv3_2023may.onnx']
861+
78.03 78.38 77.17 [640, 480] PPOCRDet with ['text_detection_en_ppocrv3_2023may.onnx']
862+
99.09 79.42 77.17 [640, 480] PPOCRDet with ['text_detection_cn_ppocrv3_2023may_int8.onnx']
863+
112.82 116.06 77.17 [640, 480] PPOCRDet with ['text_detection_en_ppocrv3_2023may_int8.onnx']
864+
142.97 142.84 135.56 [1280, 720] CRNN with ['text_recognition_CRNN_CH_2021sep.onnx']
865+
144.53 148.52 135.56 [1280, 720] CRNN with ['text_recognition_CRNN_CN_2021nov.onnx']
866+
134.47 146.62 112.91 [1280, 720] CRNN with ['text_recognition_CRNN_EN_2021sep.onnx']
867+
136.37 131.39 112.91 [1280, 720] CRNN with ['text_recognition_CRNN_CH_2023feb_fp16.onnx']
868+
132.08 117.15 109.24 [1280, 720] CRNN with ['text_recognition_CRNN_EN_2023feb_fp16.onnx']
869+
135.17 130.23 109.24 [1280, 720] CRNN with ['text_recognition_CRNN_CH_2022oct_int8.onnx']
870+
138.38 143.25 109.24 [1280, 720] CRNN with ['text_recognition_CRNN_CN_2021nov_int8.onnx']
871+
137.08 134.22 109.24 [1280, 720] CRNN with ['text_recognition_CRNN_EN_2022oct_int8.onnx']
824872
```
825873
826-
CUDA:
874+
GPU (CUDA-FP32):
827875
828876
```
829-
13.69 13.69 13.04 [1280, 720] VitTrack with ['object_tracking_vittrack_2023sep.onnx']
877+
$ python3 benchmark.py --all --fp32 --cfg_exclude wechat --cfg_overwrite_backend_target 1
878+
Benchmarking ...
879+
backend=cv.dnn.DNN_BACKEND_CUDA
880+
target=cv.dnn.DNN_TARGET_CUDA
881+
mean median min input size model
882+
5.23 5.27 5.17 [160, 120] YuNet with ['face_detection_yunet_2023mar.onnx']
883+
7.59 7.62 7.55 [150, 150] SFace with ['face_recognition_sface_2021dec.onnx']
884+
8.48 8.46 8.37 [112, 112] FacialExpressionRecog with ['facial_expression_recognition_mobilefacenet_2022july.onnx']
885+
12.29 13.04 11.11 [224, 224] MPHandPose with ['handpose_estimation_mediapipe_2023feb.onnx']
886+
12.91 13.28 12.79 [192, 192] PPHumanSeg with ['human_segmentation_pphumanseg_2023mar.onnx']
887+
8.41 8.42 8.35 [224, 224] MobileNet with ['image_classification_mobilenetv1_2022apr.onnx']
888+
9.36 9.43 8.35 [224, 224] MobileNet with ['image_classification_mobilenetv2_2022apr.onnx']
889+
32.58 32.71 31.11 [224, 224] PPResNet with ['image_classification_ppresnet50_2022jan.onnx']
890+
16.33 16.08 16.04 [320, 240] LPD_YuNet with ['license_plate_detection_lpd_yunet_2023mar.onnx']
891+
24.46 24.35 24.01 [416, 416] NanoDet with ['object_detection_nanodet_2022nov.onnx']
892+
103.28 103.41 102.37 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
893+
19.75 19.78 19.10 [1280, 720] VitTrack with ['object_tracking_vittrack_2023sep.onnx']
894+
10.84 10.76 10.75 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
895+
14.50 14.50 14.36 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
896+
23.53 23.36 23.16 [128, 256] YoutuReID with ['person_reid_youtu_2021nov.onnx']
897+
26.54 27.22 25.99 [256, 256] MPPose with ['pose_estimation_mediapipe_2023mar.onnx']
898+
27.49 27.80 26.97 [640, 480] PPOCRDet with ['text_detection_cn_ppocrv3_2023may.onnx']
899+
27.53 27.75 26.95 [640, 480] PPOCRDet with ['text_detection_en_ppocrv3_2023may.onnx']
900+
15.66 16.30 15.41 [1280, 720] CRNN with ['text_recognition_CRNN_CH_2021sep.onnx']
901+
15.91 15.80 15.41 [1280, 720] CRNN with ['text_recognition_CRNN_CN_2021nov.onnx']
902+
13.58 16.70 9.48 [1280, 720] CRNN with ['text_recognition_CRNN_EN_2021sep.onnx']
830903
```
831904
832-
CUDA-FP16:
905+
GPU (CUDA-FP16):
833906
834907
```
835-
16.29 15.77 15.77 [1280, 720] VitTrack with ['object_tracking_vittrack_2023sep.onnx']
908+
$ python3 benchmark.py --all --fp32 --cfg_exclude wechat --cfg_overwrite_backend_target 2
909+
Benchmarking ...
910+
backend=cv.dnn.DNN_BACKEND_CUDA
911+
target=cv.dnn.DNN_TARGET_CUDA_FP16
912+
mean median min input size model
913+
5.00 5.04 4.92 [160, 120] YuNet with ['face_detection_yunet_2023mar.onnx']
914+
5.09 5.08 5.05 [150, 150] SFace with ['face_recognition_sface_2021dec.onnx']
915+
6.81 6.86 6.66 [112, 112] FacialExpressionRecog with ['facial_expression_recognition_mobilefacenet_2022july.onnx']
916+
9.19 10.18 9.06 [224, 224] MPHandPose with ['handpose_estimation_mediapipe_2023feb.onnx']
917+
16.20 16.62 15.93 [192, 192] PPHumanSeg with ['human_segmentation_pphumanseg_2023mar.onnx']
918+
6.84 6.82 6.80 [224, 224] MobileNet with ['image_classification_mobilenetv1_2022apr.onnx']
919+
7.46 7.87 6.80 [224, 224] MobileNet with ['image_classification_mobilenetv2_2022apr.onnx']
920+
14.18 14.16 14.03 [224, 224] PPResNet with ['image_classification_ppresnet50_2022jan.onnx']
921+
13.35 13.10 13.04 [320, 240] LPD_YuNet with ['license_plate_detection_lpd_yunet_2023mar.onnx']
922+
19.94 19.95 19.50 [416, 416] NanoDet with ['object_detection_nanodet_2022nov.onnx']
923+
72.25 72.91 70.99 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
924+
22.37 22.44 21.60 [1280, 720] VitTrack with ['object_tracking_vittrack_2023sep.onnx']
925+
8.92 8.92 8.84 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
926+
11.11 11.13 10.98 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
927+
13.22 13.23 13.12 [128, 256] YoutuReID with ['person_reid_youtu_2021nov.onnx']
928+
26.79 27.04 26.24 [256, 256] MPPose with ['pose_estimation_mediapipe_2023mar.onnx']
929+
19.71 19.75 19.47 [640, 480] PPOCRDet with ['text_detection_cn_ppocrv3_2023may.onnx']
930+
19.76 19.93 19.47 [640, 480] PPOCRDet with ['text_detection_en_ppocrv3_2023may.onnx']
931+
16.30 15.88 15.80 [1280, 720] CRNN with ['text_recognition_CRNN_CH_2021sep.onnx']
932+
16.36 16.51 15.80 [1280, 720] CRNN with ['text_recognition_CRNN_CN_2021nov.onnx']
933+
13.64 16.27 8.90 [1280, 720] CRNN with ['text_recognition_CRNN_EN_2021sep.onnx']
836934
```
837935
936+
<!--
937+
838938
### Atlas 200I DK
839939
840940
CPU:

0 commit comments

Comments
 (0)