Skip to content

Commit

Permalink
Add/yolo11 seg and pose (#153)
Browse files Browse the repository at this point in the history 
* add yolo11 seg and pose
  • Loading branch information
@BUG1989
BUG1989 authored Nov 3, 2024
1 parent d885982 commit 52f1a12
Show file tree
Hide file tree
Showing 17 changed files with 704 additions and 111 deletions.
Binary file added docs/ax650/yolo11_pose_out.jpg
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added docs/ax650/yolo11_seg_out.jpg
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added docs/ax650/yolov8_pose_out.jpg
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified docs/ax650/yolov8_seg_out.jpg
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file removed docs/ax650/yolov8s_pose_out.jpg
View file
Binary file not shown.
4 changes: 2 additions & 2 deletions examples/ax620e/CMakeLists.txt
View file
Original file line number Diff line number Diff line change
Expand Up @@ -28,9 +28,9 @@ axera_example(ax_yolov5_face ax_yolov5_face_steps.cc)
axera_example(ax_yolov6 ax_yolov6_steps.cc)
axera_example(ax_yolov7_tiny_face ax_yolov7_tiny_face_steps.cc)
axera_example(ax_yolov7 ax_yolov7_steps.cc)
axera_example(ax_yolov8 ax_yolov8s_steps.cc)
axera_example(ax_yolov8 ax_yolov8_steps.cc)
axera_example(ax_yolov8_seg ax_yolov8_seg_steps.cc)
axera_example(ax_yolov8_pose ax_yolov8s_pose_steps.cc)
axera_example(ax_yolov8_pose ax_yolov8_pose_steps.cc)
axera_example(ax_yolov9 ax_yolov9_steps.cc)
axera_example(ax_yolov10 ax_yolov10s_steps.cc)
axera_example(ax_yolov10_u ax_yolov10_ultralytics_steps.cc)
Expand Down
16 changes: 12 additions & 4 deletions examples/ax620e/ax_yolov8s_pose_steps.cc → examples/ax620e/ax_yolov8_pose_steps.cc
View file
Original file line number Diff line number Diff line change
Expand Up @@ -59,14 +59,22 @@ namespace ax
std::vector<detection::Object> proposals;
std::vector<detection::Object> objects;
timer timer_postprocess;

float* output_ptr[3] = {(float*)io_data->pOutputs[0].pVirAddr, // 1*80*80*65
(float*)io_data->pOutputs[1].pVirAddr, // 1*40*40*65
(float*)io_data->pOutputs[2].pVirAddr}; // 1*20*20*65
float* output_kps_ptr[3] = {(float*)io_data->pOutputs[3].pVirAddr, // 1*80*80*51
(float*)io_data->pOutputs[4].pVirAddr, // 1*40*40*51
(float*)io_data->pOutputs[5].pVirAddr}; // 1*20*20*51

for (int i = 0; i < 3; ++i)
{
auto feat_ptr = (float*)io_data->pOutputs[i + 3].pVirAddr;
auto feat_kps_ptr = (float*)io_data->pOutputs[i].pVirAddr;
auto feat_ptr = output_ptr[i];
auto feat_kps_ptr = output_kps_ptr[i];
int32_t stride = (1 << i) * 8;
detection::generate_proposals_yolov8_pose_native(stride, feat_ptr, feat_kps_ptr, PROB_THRESHOLD, proposals, input_w, input_h, NUM_POINT, NUM_CLASS);
}

detection::get_out_bbox_kps(proposals, objects, NMS_THRESHOLD, input_h, input_w, mat.rows, mat.cols);
fprintf(stdout, "post process cost time:%.2f ms \n", timer_postprocess.cost());
fprintf(stdout, "--------------------------------------\n");
Expand All @@ -81,7 +89,7 @@ namespace ax
fprintf(stdout, "--------------------------------------\n");
fprintf(stdout, "detection num: %zu\n", objects.size());

detection::draw_keypoints(mat, objects, KPS_COLORS, LIMB_COLORS, SKELETON, "yolov8s_pose_out");
detection::draw_keypoints(mat, objects, KPS_COLORS, LIMB_COLORS, SKELETON, "yolov8_pose_out");
}

bool run_model(const std::string& model, const std::vector<uint8_t>& data, const int& repeat, cv::Mat& mat, int input_h, int input_w)
Expand Down
0 examples/ax620e/ax_yolov8s_steps.cc → examples/ax620e/ax_yolov8_steps.cc
View file
File renamed without changes.
2 changes: 2 additions & 0 deletions examples/ax650/CMakeLists.txt
View file
Original file line number Diff line number Diff line change
Expand Up @@ -39,6 +39,8 @@ axera_example(ax_yolox ax_yolox_steps.cc)
axera_example(ax_yolov10 ax_yolov10s_steps.cc)
axera_example(ax_yolov10_u ax_yolov10_ultralytics_steps.cc)
axera_example(ax_yolo11 ax_yolo11_steps.cc)
axera_example(ax_yolo11_seg ax_yolo11_seg_steps.cc)
axera_example(ax_yolo11_pose ax_yolo11_pose_steps.cc)
axera_example(ax_yolo_nas ax_yolo_nas_steps.cc)
axera_example(ax_yolo_world ax_yolo_world_steps.cc)
axera_example(ax_yolo_world_open_vocabulary ax_yolo_world_open_vocabulary_steps.cc)
Expand Down
251 changes: 166 additions & 85 deletions examples/ax650/README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,8 @@ AX-Samples 将不断更新最流行的、实用的、有趣的示例代码。
- [YOLO11](#YOLO11)
- 物体分割
- [YOLOv5-seg](#YOLOv5-seg)
- [YOLOv8-seg](#YOLOv8-seg)
- [YOLO11-seg](#YOLO11-seg)
- 人脸检测
- [scrfd](#Scrfd)
- [YOLOv5-Face](#YOLOv5-Face)([original model](https://github.com/deepcam-cn/yolov5-face))
Expand All @@ -21,6 +23,8 @@ AX-Samples 将不断更新最流行的、实用的、有趣的示例代码。
- [YOLOv5s_visdrone](#YOLOv5s_visdrone)
- 人体关键点
- [HRNet](#HRNet)
- [YOLOv8-pose](#YOLOv8-pose)
- [YOLO11-pose](#YOLO11-pose)
- 人体分割
- [PP-HumanSeg](#PP-HumanSeg)

Expand Down Expand Up @@ -331,6 +335,87 @@ detection num: 11
```
![YOLOv8s](../../docs/ax650/yolov8s_out.jpg)

#### YOLOv8-seg
```
root@ax650:/# .ax_yolov8_seg -m yolov8s_seg.axmodel -i ssd_car.jpg
--------------------------------------
model file : yolov8s_seg.axmodel
image file : ssd_car.jpg
img_h, img_w : 640 640
--------------------------------------
Engine creating handle is done.
Engine creating context is done.
Engine get io info is done.
Engine alloc io is done.
Engine push input is done.
--------------------------------------
input size: 1
name: images [UINT8] [BGR]
1 x 640 x 640 x 3
output size: 7
name: /model.22/Concat_1_output_0 [FLOAT32]
1 x 80 x 80 x 144
name: /model.22/Concat_2_output_0 [FLOAT32]
1 x 40 x 40 x 144
name: /model.22/Concat_3_output_0 [FLOAT32]
1 x 20 x 20 x 144
name: /model.22/cv4.0/cv4.0.2/Conv_output_0 [FLOAT32]
1 x 80 x 80 x 32
name: /model.22/cv4.1/cv4.1.2/Conv_output_0 [FLOAT32]
1 x 40 x 40 x 32
name: /model.22/cv4.2/cv4.2.2/Conv_output_0 [FLOAT32]
1 x 20 x 20 x 32
name: output1 [FLOAT32]
1 x 32 x 160 x 160
post process cost time:8.68 ms
--------------------------------------
Repeat 1 times, avg time 4.69 ms, max_time 4.69 ms, min_time 4.69 ms
--------------------------------------
detection num: 3
2: 96%, [ 330, 202, 499, 326], car
0: 91%, [ 205, 185, 286, 373], person
5: 89%, [ 128, 67, 450, 299], bus
--------------------------------------
```
![YOLOv8-seg](../../docs/ax650/yolov8_seg_out.jpg)

#### YOLOv8-pose
```
root@ax650:/# ./ax_yolov8_pose -m yolov8s_pose.axmodel -i pose_test.jpg
--------------------------------------
model file : yolov8s_pose.axmodel
image file : pose_test.jpg
img_h, img_w : 640 640
--------------------------------------
Engine creating handle is done.
Engine creating context is done.
Engine get io info is done.
Engine alloc io is done.
Engine push input is done.
--------------------------------------
post process cost time:1.26 ms
--------------------------------------
Repeat 1 times, avg time 3.74 ms, max_time 3.74 ms, min_time 3.74 ms
--------------------------------------
detection num: 4
0: 93%, [ 756, 212, 1127, 1158], person
0: 91%, [ 0, 358, 316, 1105], person
0: 91%, [1349, 338, 1629, 1034], person
0: 87%, [ 489, 474, 656, 996], person
--------------------------------------
```
![YOLOv8-pose](../../docs/ax650/yolov8_pose_out.jpg)

#### YOLOX-S
```
root@AXERA:/home/test# ./ax_yolox -m yolox.axmodel -i test.jpg
Expand Down Expand Up @@ -466,35 +551,6 @@ Repeat 100 times, avg time 0.44 ms, max_time 0.45 ms, min_time 0.44 ms
```
![PFLD](../../docs/ax650/pfld_out.jpg)

### YOLOV8-POSE
```
/opt/test # ./ax_yolov8_pose -m yolov8s-pose.axmodel -i test.jpg
--------------------------------------
model file : yolov8s-pose.axmodel
image file : test.jpg
img_h, img_w : 640 640
--------------------------------------
[Axera version]: libax_sys.so V1.14.0_20230506154237 May 6 2023 15:43:14 JK
Engine creating handle is done.
Engine creating context is done.
Engine get io info is done.
Engine alloc io is done.
Engine push input is done.
--------------------------------------
post process cost time:0.28 ms
--------------------------------------
Repeat 1 times, avg time 12.77 ms, max_time 12.77 ms, min_time 12.77 ms
--------------------------------------
detection num: 5
0: 92%, [ 497, 291, 673, 804], person
0: 90%, [ 175, 306, 365, 804], person
0: 86%, [ 89, 292, 192, 490], person
0: 77%, [ 608, 271, 693, 465], person
0: 55%, [ 443, 296, 498, 440], person
--------------------------------------
```
![yolov8-pose](../../docs/ax650/yolov8s_pose_out.jpg)

### DinoV2
```
/opt/test # ./ax_dinov2 -m dinov2_small_518_precision_opt.axmodel -i dog-chai.jpeg
Expand Down Expand Up @@ -583,62 +639,6 @@ Repeat 1 times, avg time 56.67 ms, max_time 56.67 ms, min_time 56.67 ms
```
<img src="../../docs/ax650/depth_anything_out.png">

### YoloV8-Seg
```
/opt/test # ./ax_yolov8_seg -m yolov8n_seg.axmodel -i ssd_horse.jpg
--------------------------------------
model file : yolov8n_seg.axmodel
image file : ssd_horse.jpg
img_h, img_w : 640 640
--------------------------------------
Engine creating handle is done.
Engine creating context is done.
Engine get io info is done.
Engine alloc io is done.
Engine push input is done.
--------------------------------------
input size: 1
name: images [UINT8]
1 x 640 x 640 x 3
output size: 7
name: output1 [FLOAT32]
1 x 32 x 160 x 160
name: /model.22/cv4.0/cv4.0.2/Conv_output_0 [FLOAT32]
1 x 80 x 80 x 32
name: /model.22/cv4.1/cv4.1.2/Conv_output_0 [FLOAT32]
1 x 40 x 40 x 32
name: /model.22/cv4.2/cv4.2.2/Conv_output_0 [FLOAT32]
1 x 20 x 20 x 32
name: /model.22/Concat_1_output_0 [FLOAT32]
1 x 80 x 80 x 144
name: /model.22/Concat_2_output_0 [FLOAT32]
1 x 40 x 40 x 144
name: /model.22/Concat_3_output_0 [FLOAT32]
1 x 20 x 20 x 144
post process cost time:8.62 ms
--------------------------------------
Repeat 1 times, avg time 5.16 ms, max_time 5.16 ms, min_time 5.16 ms
--------------------------------------
detection num: 5
17: 90%, [ 214, 75, 419, 369], horse
0: 83%, [ 272, 13, 349, 232], person
7: 77%, [ 0, 105, 133, 196], truck
0: 77%, [ 427, 125, 451, 175], person
16: 69%, [ 144, 203, 195, 343], dog
--------------------------------------
```
<img src="../../docs/ax650/yolov8_seg_out.jpg">

### YOLOV8-OBB
```
/opt/test # ./ax_yolov8_obb -m ./yolov8s-obb.axmodel -i ./dota_demo.jpg -r 10
Expand Down Expand Up @@ -779,6 +779,87 @@ detection num: 6
```
![](../../docs/ax650/yolo11_out.jpg)

#### YOLO11-seg
```
root@ax650:/# ./ax_yolo11_seg -m yolo11s_seg.axmodel -i ssd_car.jpg
--------------------------------------
model file : yolo11s_seg.axmodel
image file : ssd_car.jpg
img_h, img_w : 640 640
--------------------------------------
Engine creating handle is done.
Engine creating context is done.
Engine get io info is done.
Engine alloc io is done.
Engine push input is done.
--------------------------------------
input size: 1
name: images [UINT8] [BGR]
1 x 640 x 640 x 3
output size: 7
name: /model.22/Concat_1_output_0 [FLOAT32]
1 x 80 x 80 x 144
name: /model.22/Concat_2_output_0 [FLOAT32]
1 x 40 x 40 x 144
name: /model.22/Concat_3_output_0 [FLOAT32]
1 x 20 x 20 x 144
name: /model.22/cv4.0/cv4.0.2/Conv_output_0 [FLOAT32]
1 x 80 x 80 x 32
name: /model.22/cv4.1/cv4.1.2/Conv_output_0 [FLOAT32]
1 x 40 x 40 x 32
name: /model.22/cv4.2/cv4.2.2/Conv_output_0 [FLOAT32]
1 x 20 x 20 x 32
name: output1 [FLOAT32]
1 x 32 x 160 x 160
post process cost time:8.68 ms
--------------------------------------
Repeat 1 times, avg time 4.69 ms, max_time 4.69 ms, min_time 4.69 ms
--------------------------------------
detection num: 3
2: 96%, [ 330, 202, 499, 326], car
0: 91%, [ 205, 185, 286, 373], person
5: 89%, [ 128, 67, 450, 299], bus
--------------------------------------
```
![YOLO11-seg](../../docs/ax650/yolo11_seg_out.jpg)

#### YOLO11-pose
```
root@ax650:/# ./ax_yolo11_pose -m yolo11s_pose.axmodel -i pose_test.jpg
--------------------------------------
model file : yolo11s_pose.axmodel
image file : pose_test.jpg
img_h, img_w : 640 640
--------------------------------------
Engine creating handle is done.
Engine creating context is done.
Engine get io info is done.
Engine alloc io is done.
Engine push input is done.
--------------------------------------
post process cost time:1.31 ms
--------------------------------------
Repeat 1 times, avg time 3.35 ms, max_time 3.35 ms, min_time 3.35 ms
--------------------------------------
detection num: 4
0: 94%, [ 761, 220, 1128, 1153], person
0: 91%, [1352, 343, 1633, 1033], person
0: 89%, [ 488, 477, 661, 996], person
0: 81%, [ 0, 357, 317, 1110], person
--------------------------------------
```
![YOLO11-pose](../../docs/ax650/yolo11_pose_out.jpg)

### YOLO-WORLD-OPEN-VOCABULARY
```
root@ax650: ~# ./ax_yolo_world_open_vocabulary -m yoloworld.axmodel -t data/person.bin -i ../../images/ssd_horse.jpg
Expand Down
Loading

0 comments on commit 52f1a12

Please sign in to comment.