Skip to content

Commit b0fe213

Browse files
fengyuentaufengyuentau
authored
Remove DaSiamRPN since we have its superseder VitTrack now (#213)
* first commit * fix path
1 parent fef72f8 commit b0fe213

15 files changed

+152
-783
lines changed

‎README.md

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -71,9 +71,9 @@ Some examples are listed below. You can find more in the directory of each model
7171

7272
![yolox demo](./models/object_detection_yolox/example_outputs/3_res.jpg)
7373

74-
### Object Tracking with [DaSiamRPN](./models/object_tracking_dasiamrpn/)
74+
### Object Tracking with [VitTrack](./models/object_tracking_vittrack/)
7575

76-
![webcam demo](./models/object_tracking_dasiamrpn/example_outputs/dasiamrpn_demo.gif)
76+
![webcam demo](./models/object_tracking_vittrack/example_outputs/vittrack_demo.gif)
7777

7878
### Palm Detection with [MP-PalmDet](./models/palm_detection_mediapipe/)
7979

‎benchmark/README.md

Lines changed: 4 additions & 15 deletions
Original file line numberDiff line numberDiff line change
@@ -31,7 +31,7 @@ python benchmark.py --all --fp32
3131

3232
# All configs but exclude some of them (fill with config name keywords, not sensitive to upper/lower case, seperate with colons)
3333
python benchmark.py --all --cfg_exclude wechat
34-
python benchmark.py --all --cfg_exclude wechat:dasiamrpn
34+
python benchmark.py --all --cfg_exclude wechat:crnn
3535

3636
# All configs but exclude some of the models (fill with exact model names, sensitive to upper/lower case, seperate with colons)
3737
python benchmark.py --all --model_exclude license_plate_detection_lpd_yunet_2023mar_int8.onnx:human_segmentation_pphumanseg_2023mar_int8.onnx
@@ -94,7 +94,6 @@ mean median min input size model
9494
46.10 47.53 43.06 [416, 416] NanoDet with ['object_detection_nanodet_2022nov_int8.onnx']
9595
144.89 149.58 125.71 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
9696
143.83 146.39 119.75 [640, 640] YoloX with ['object_detection_yolox_2022nov_int8.onnx']
97-
23.43 22.82 20.90 [1280, 720] DaSiamRPN with ['object_tracking_dasiamrpn_kernel_cls1_2021nov.onnx', 'object_tracking_dasiamrpn_kernel_r1_2021nov.onnx', 'object_tracking_dasiamrpn_model_2021nov.onnx']
9897
12.99 13.11 12.14 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
9998
12.64 12.44 10.82 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb_int8.onnx']
10099
12.64 11.83 11.03 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
@@ -149,7 +148,6 @@ mean median min input size model
149148
212.90 212.93 209.55 [416, 416] NanoDet with ['object_detection_nanodet_2022nov_int8.onnx']
150149
1690.06 2303.34 1480.63 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
151150
1489.54 1435.48 1308.12 [640, 640] YoloX with ['object_detection_yolox_2022nov_int8.onnx']
152-
564.90 580.35 527.49 [1280, 720] DaSiamRPN with ['object_tracking_dasiamrpn_kernel_cls1_2021nov.onnx', 'object_tracking_dasiamrpn_kernel_r1_2021nov.onnx', 'object_tracking_dasiamrpn_model_2021nov.onnx']
153151
356.63 357.29 354.42 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
154152
217.52 229.39 101.61 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb_int8.onnx']
155153
198.63 198.25 196.68 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
@@ -205,7 +203,6 @@ mean median min input size model
205203
216.18 216.19 214.30 [416, 416] NanoDet with ['object_detection_nanodet_2022nov_int8.onnx']
206204
1207.83 1208.71 1203.64 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
207205
1236.98 1250.21 1203.64 [640, 640] YoloX with ['object_detection_yolox_2022nov_int8.onnx']
208-
456.79 456.90 445.83 [1280, 720] DaSiamRPN with ['object_tracking_dasiamrpn_kernel_cls1_2021nov.onnx', 'object_tracking_dasiamrpn_kernel_r1_2021nov.onnx', 'object_tracking_dasiamrpn_model_2021nov.onnx']
209206
124.89 125.25 124.53 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
210207
107.99 109.82 94.05 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb_int8.onnx']
211208
108.41 108.33 107.91 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
@@ -244,7 +241,6 @@ mean median min input size model
244241
54.24 55.24 52.87 [320, 240] LPD_YuNet with ['license_plate_detection_lpd_yunet_2023mar.onnx']
245242
63.63 63.43 63.32 [416, 416] NanoDet with ['object_detection_nanodet_2022nov.onnx']
246243
371.45 378.00 366.39 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
247-
77.50 77.73 76.16 [1280, 720] DaSiamRPN with ['object_tracking_dasiamrpn_kernel_cls1_2021nov.onnx', 'object_tracking_dasiamrpn_kernel_r1_2021nov.onnx', 'object_tracking_dasiamrpn_model_2021nov.onnx']
248244
33.85 33.90 33.61 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
249245
38.16 37.33 37.10 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
250246
91.65 91.98 89.90 [128, 256] YoutuReID with ['person_reid_youtu_2021nov.onnx']
@@ -275,7 +271,6 @@ mean median min input size model
275271
366.46 366.88 363.46 [320, 240] LPD_YuNet with ['license_plate_detection_lpd_yunet_2023mar.onnx']
276272
163.06 163.34 161.77 [416, 416] NanoDet with ['object_detection_nanodet_2022nov.onnx']
277273
301.10 311.52 297.74 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
278-
53.34 54.30 51.79 [1280, 720] DaSiamRPN with ['object_tracking_dasiamrpn_kernel_cls1_2021nov.onnx', 'object_tracking_dasiamrpn_kernel_r1_2021nov.onnx', 'object_tracking_dasiamrpn_model_2021nov.onnx']
279274
149.37 149.95 148.01 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
280275
153.89 153.96 153.43 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
281276
44.29 44.03 43.62 [128, 256] YoutuReID with ['person_reid_youtu_2021nov.onnx']
@@ -323,7 +318,6 @@ mean median min input size model
323318
212.69 262.75 170.88 [416, 416] NanoDet with ['object_detection_nanodet_2022nov_int8.onnx']
324319
1110.87 1112.27 1085.31 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
325320
1128.73 1157.12 1085.31 [640, 640] YoloX with ['object_detection_yolox_2022nov_int8.onnx']
326-
382.57 464.42 354.66 [1280, 720] DaSiamRPN with ['object_tracking_dasiamrpn_kernel_cls1_2021nov.onnx', 'object_tracking_dasiamrpn_kernel_r1_2021nov.onnx', 'object_tracking_dasiamrpn_model_2021nov.onnx']
327321
147.01 144.01 139.27 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
328322
119.70 118.95 94.09 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb_int8.onnx']
329323
107.63 107.09 105.61 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
@@ -404,7 +398,6 @@ mean median min input size model
404398
322.98 323.45 312.13 [416, 416] NanoDet with ['object_detection_nanodet_2022nov_int8.onnx']
405399
1875.33 1877.53 1871.26 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
406400
1989.04 2005.25 1871.26 [640, 640] YoloX with ['object_detection_yolox_2022nov_int8.onnx']
407-
637.54 640.61 626.98 [1280, 720] DaSiamRPN with ['object_tracking_dasiamrpn_kernel_cls1_2021nov.onnx', 'object_tracking_dasiamrpn_kernel_r1_2021nov.onnx', 'object_tracking_dasiamrpn_model_2021nov.onnx']
408401
159.80 159.62 159.40 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
409402
152.18 152.86 145.56 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb_int8.onnx']
410403
145.83 145.77 145.45 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
@@ -426,7 +419,7 @@ mean median min input size model
426419
NPU (CANN):
427420
428421
```
429-
$ python3 benchmark.py --all --fp32 --cfg_exclude wechat:dasiamrpn:crnn --model_exclude pose_estimation_mediapipe_2023mar.onnx --cfg_overwrite_backend_target 4
422+
$ python3 benchmark.py --all --fp32 --cfg_exclude wechat:crnn --model_exclude pose_estimation_mediapipe_2023mar.onnx --cfg_overwrite_backend_target 4
430423
Benchmarking ...
431424
backend=cv.dnn.DNN_BACKEND_CANN
432425
target=cv.dnn.DNN_TARGET_NPU
@@ -485,7 +478,6 @@ mean median min input size model
485478
1903.82 1962.71 1533.79 [416, 416] NanoDet with ['object_detection_nanodet_2022nov_int8.onnx']
486479
37604.10 37569.30 37502.48 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
487480
24229.20 25577.94 13483.54 [640, 640] YoloX with ['object_detection_yolox_2022nov_int8.onnx']
488-
14860.23 14988.15 14769.91 [1280, 720] DaSiamRPN with ['object_tracking_dasiamrpn_kernel_cls1_2021nov.onnx', 'object_tracking_dasiamrpn_kernel_r1_2021nov.onnx', 'object_tracking_dasiamrpn_model_2021nov.onnx']
489481
1133.44 1131.54 1124.83 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
490482
883.96 919.07 655.33 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb_int8.onnx']
491483
1430.98 1424.55 1415.68 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
@@ -541,7 +533,6 @@ mean median min input size model
541533
117.28 150.31 83.33 [416, 416] NanoDet with ['object_detection_nanodet_2022nov_int8.onnx']
542534
553.58 558.76 535.47 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
543535
594.18 592.64 535.47 [640, 640] YoloX with ['object_detection_yolox_2022nov_int8.onnx']
544-
138.82 151.00 113.82 [1280, 720] DaSiamRPN with ['object_tracking_dasiamrpn_kernel_cls1_2021nov.onnx', 'object_tracking_dasiamrpn_kernel_r1_2021nov.onnx', 'object_tracking_dasiamrpn_model_2021nov.onnx']
545536
56.35 55.73 55.25 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
546537
57.07 57.19 55.25 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb_int8.onnx']
547538
47.94 48.41 47.05 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
@@ -596,7 +587,6 @@ mean median min input size model
596587
406.28 416.58 385.68 [416, 416] NanoDet with ['object_detection_nanodet_2022nov_int8.onnx']
597588
2608.90 2612.42 2597.93 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
598589
2609.88 2609.39 2597.93 [640, 640] YoloX with ['object_detection_yolox_2022nov_int8.onnx']
599-
809.55 814.66 794.67 [1280, 720] DaSiamRPN with ['object_tracking_dasiamrpn_kernel_cls1_2021nov.onnx', 'object_tracking_dasiamrpn_kernel_r1_2021nov.onnx', 'object_tracking_dasiamrpn_model_2021nov.onnx']
600590
228.95 228.74 228.35 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
601591
227.97 228.61 226.76 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb_int8.onnx']
602592
192.29 192.26 191.74 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
@@ -653,7 +643,6 @@ mean median min input size model
653643
3002.36 3047.94 2655.38 [416, 416] NanoDet with ['object_detection_nanodet_2022nov_int8.onnx']
654644
50678.08 50651.82 50651.19 [640, 640] YoloX with ['object_detection_yolox_2022nov.onnx']
655645
36249.71 37771.22 24606.37 [640, 640] YoloX with ['object_detection_yolox_2022nov_int8.onnx']
656-
19974.99 19984.80 19948.63 [1280, 720] DaSiamRPN with ['object_tracking_dasiamrpn_kernel_cls1_2021nov.onnx', 'object_tracking_dasiamrpn_kernel_r1_2021nov.onnx', 'object_tracking_dasiamrpn_model_2021nov.onnx']
657646
1502.15 1501.98 1500.99 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb.onnx']
658647
1300.15 1320.44 1137.60 [192, 192] MPPalmDet with ['palm_detection_mediapipe_2023feb_int8.onnx']
659648
1993.05 1993.98 1991.86 [224, 224] MPPersonDet with ['person_detection_mediapipe_2023mar.onnx']
@@ -680,9 +669,9 @@ Specs: [details_cn](https://doc.rvspace.org/VisionFive2/PB/VisionFive_2/specific
680669
681670
CPU:
682671
<!-- config wechat is excluded due to it needs building with opencv_contrib -->
683-
<!-- config dasiam is excluded due to opencv cannot find ffmpeg and its components -->
672+
<!-- config vittrack is excluded due to opencv cannot find ffmpeg and its components -->
684673
```
685-
$ python3 benchmark.py --all --cfg_exclude wechat:dasiam
674+
$ python3 benchmark.py --all --cfg_exclude wechat:vittrack
686675
Benchmarking ...
687676
backend=cv.dnn.DNN_BACKEND_OPENCV
688677
target=cv.dnn.DNN_TARGET_CPU

0 commit comments

Comments
 (0)