@@ -101,7 +101,7 @@ PP-YOLO improved performance and speed of YOLOv3 with following methods:
...
@@ -101,7 +101,7 @@ PP-YOLO improved performance and speed of YOLOv3 with following methods:
**Notes:**
**Notes:**
- PP-YOLO-tiny is trained on COCO train2017 datast and evaluated on val2017 dataset，Box AP<sup>val</sup> is evaluation results of `mAP(IoU=0.5:0.95)`, Box AP<sup>val</sup> is evaluation results of `mAP(IoU=0.5)`.
- PP-YOLO-tiny is trained on COCO train2017 datast and evaluated on val2017 dataset，Box AP<sup>val</sup> is evaluation results of `mAP(IoU=0.5:0.95)`, Box AP<sup>val</sup> is evaluation results of `mAP(IoU=0.5)`.
- PP-YOLO-tiny used 8 GPUs for training and mini-batch size as 32 on each GPU, if GPU number and mini-batch size is changed, learning rate and iteration times should be adjusted according [FAQ](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/FAQ.md).
- PP-YOLO-tiny used 8 GPUs for training and mini-batch size as 32 on each GPU, if GPU number and mini-batch size is changed, learning rate and iteration times should be adjusted according [FAQ](https://github.com/PaddlePaddle/PaddleDetection/blob//develop/static/docs/FAQ.md).
- PP-YOLO-tiny inference speed is tested on Kirin 990 with 4 threads by arm8
- PP-YOLO-tiny inference speed is tested on Kirin 990 with 4 threads by arm8
- we alse provide PP-YOLO-tiny post quant inference model, which can compress model to **1.3MB** with nearly no inference on inference speed and performance
- we alse provide PP-YOLO-tiny post quant inference model, which can compress model to **1.3MB** with nearly no inference on inference speed and performance
**Attention**: Pretrained models whose congigurations are in the directory `generic` just support inference but do not support training and evaluation as now.
**Attention**: Pretrained models whose congigurations are in the directory `generic` just support inference but do not support training and evaluation as now.
- Deformable ConvNets v2(dcn_v2) reference from [Deformable ConvNets v2](https://arxiv.org/abs/1811.11168).
- Deformable ConvNets v2(dcn_v2) reference from [Deformable ConvNets v2](https://arxiv.org/abs/1811.11168).
-`c3-c5` means adding `dcn` in resnet stage 3 to 5.
-`c3-c5` means adding `dcn` in resnet stage 3 to 5.
- Detailed configuration file in [configs/dcn](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/dcn)
- Detailed configuration file in [configs/dcn](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/dcn)
### HRNet
### HRNet
* See more details in [HRNet model zoo](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/hrnet/).
* See more details in [HRNet model zoo](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/hrnet/).
### Res2Net
### Res2Net
* See more details in [Res2Net model zoo](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/res2net/).
* See more details in [Res2Net model zoo](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/res2net/).
### IOU loss
### IOU loss
* GIOU loss and DIOU loss are included now. See more details in [IOU loss model zoo](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/iou_loss/).
* GIOU loss and DIOU loss are included now. See more details in [IOU loss model zoo](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/iou_loss/).
### GCNet
### GCNet
* See more details in [GCNet model zoo](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/gcnet/).
* See more details in [GCNet model zoo](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/gcnet/).
### Libra R-CNN
### Libra R-CNN
* See more details in [Libra R-CNN model zoo](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/libra_rcnn/).
* See more details in [Libra R-CNN model zoo](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/libra_rcnn/).
### Auto Augmentation
### Auto Augmentation
* See more details in [Auto Augmentation model zoo](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/autoaugment/).
* See more details in [Auto Augmentation model zoo](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/autoaugment/).
### Group Normalization
### Group Normalization
| Backbone | Type | Image/gpu | Lr schd | Box AP | Mask AP | Download | Configs |
| Backbone | Type | Image/gpu | Lr schd | Box AP | Mask AP | Download | Configs |
**NOTE**: MobileNet-SSD is trained in 2 GPU with totoal batch size as 64 and trained 120 epoches. VGG-SSD is trained in 4 GPU with total batch size as 32 and trained 240 epoches. SSD training data augmentations: randomly color distortion,
**NOTE**: MobileNet-SSD is trained in 2 GPU with totoal batch size as 64 and trained 120 epoches. VGG-SSD is trained in 4 GPU with total batch size as 32 and trained 240 epoches. SSD training data augmentations: randomly color distortion,
@@ -11,7 +11,7 @@ In transfer learning, if different dataset and the number of classes is used, th
...
@@ -11,7 +11,7 @@ In transfer learning, if different dataset and the number of classes is used, th
Transfer learning needs custom dataset and annotation in COCO-format and VOC-format is supported now. The script converts the annotation from voc, labelme or cityscape to COCO is provided in ```tools/x2coco.py```. More details please refer to [READER](READER.md). After data preparation, update the data parameters in configuration file.
Transfer learning needs custom dataset and annotation in COCO-format and VOC-format is supported now. The script converts the annotation from voc, labelme or cityscape to COCO is provided in ```tools/x2coco.py```. More details please refer to [READER](READER.md). After data preparation, update the data parameters in configuration file.
1. COCO-format dataset, take [yolov3\_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet.yml#L66) for example, modify the COCODataSet in yolov3\_reader:
1. COCO-format dataset, take [yolov3\_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet.yml#L66) for example, modify the COCODataSet in yolov3\_reader:
```yml
```yml
dataset:
dataset:
...
@@ -22,7 +22,7 @@ Transfer learning needs custom dataset and annotation in COCO-format and VOC-for
...
@@ -22,7 +22,7 @@ Transfer learning needs custom dataset and annotation in COCO-format and VOC-for
with_background:false
with_background:false
```
```
2. VOC-format dataset, take [yolov3\_darknet\_voc.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet_voc.yml#L67) for example, modify the VOCDataSet in the configuration:
2. VOC-format dataset, take [yolov3\_darknet\_voc.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet_voc.yml#L67) for example, modify the VOCDataSet in the configuration:
The parameters which need to ignore can be specified explicitly as well and arbitrary parameter names can be added to `finetune_exclude_pretrained_params`. For this purpose, several methods can be used as follwed:
The parameters which need to ignore can be specified explicitly as well and arbitrary parameter names can be added to `finetune_exclude_pretrained_params`. For this purpose, several methods can be used as follwed:
- Set `finetune_exclude_pretrained_params` in YAML configuration files. Please refer to [configure file](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_mobilenet_v1_fruit.yml#L15)
- Set `finetune_exclude_pretrained_params` in YAML configuration files. Please refer to [configure file](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_mobilenet_v1_fruit.yml#L15)
- Set `finetune_exclude_pretrained_params` in command line. For example:
- Set `finetune_exclude_pretrained_params` in command line. For example:
本文档步骤，我们同时在`Visual Studio 2015` 和 `Visual Studio 2019 Community` 两个版本进行了测试，我们推荐使用[`Visual Studio 2019`直接编译`CMake`项目](https://github.com/PaddlePaddle/PaddleDetection/blob/master/deploy/cpp/docs/windows_vs2019_build.md)。
本文档步骤，我们同时在`Visual Studio 2015` 和 `Visual Studio 2019 Community` 两个版本进行了测试，我们推荐使用[`Visual Studio 2019`直接编译`CMake`项目](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/deploy/cpp/docs/windows_vs2019_build.md)。
@@ -18,7 +18,7 @@ The network for detecting vehicles is YOLOv3, the backbone of which is Dacknet53
...
@@ -18,7 +18,7 @@ The network for detecting vehicles is YOLOv3, the backbone of which is Dacknet53
### 2. Configuration for training
### 2. Configuration for training
PaddleDetection provides users with a configuration file [yolov3_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet.yml) to train YOLOv3 on the COCO dataset, compared with this file, we modify some parameters as followed to conduct the training for vehicle detection:
PaddleDetection provides users with a configuration file [yolov3_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet.yml) to train YOLOv3 on the COCO dataset, compared with this file, we modify some parameters as followed to conduct the training for vehicle detection:
* max_iters: 120000
* max_iters: 120000
* num_classes: 6
* num_classes: 6
...
@@ -67,7 +67,7 @@ The network for detecting vehicles is YOLOv3, the backbone of which is Dacknet53
...
@@ -67,7 +67,7 @@ The network for detecting vehicles is YOLOv3, the backbone of which is Dacknet53
### 2. Configuration for training
### 2. Configuration for training
PaddleDetection provides users with a configuration file [yolov3_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet.yml) to train YOLOv3 on the COCO dataset, compared with this file, we modify some parameters as followed to conduct the training for pedestrian detection:
PaddleDetection provides users with a configuration file [yolov3_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet.yml) to train YOLOv3 on the COCO dataset, compared with this file, we modify some parameters as followed to conduct the training for pedestrian detection:
- Get mAP in `Easy/Medium/Hard Set` by multi-scale evaluation in `tools/face_eval.py`.
- Get mAP in `Easy/Medium/Hard Set` by multi-scale evaluation in `tools/face_eval.py`.
For details can refer to [Evaluation](#Evaluate-on-the-WIDER-FACE).
For details can refer to [Evaluation](#Evaluate-on-the-WIDER-FACE).
- BlazeFace-Lite Training and Testing ues [blazeface.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/face_detection/blazeface.yml)
- BlazeFace-Lite Training and Testing ues [blazeface.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/face_detection/blazeface.yml)
Required python packages are specified in [requirements.txt](https://github.com/PaddlePaddle/PaddleDetection/blob/master/requirements.txt), and can be installed with:
Required python packages are specified in [requirements.txt](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/requirements.txt), and can be installed with:
