update sniper docs (#4412)

configs/sniper/README.md

@@ -6,8 +6,9 @@ English | [简体中文](README_cn.md)
 
 | Sniper   | GPU number    | images/GPU |    Model  |    Dataset     | Schedulers | Box AP |          Download                  | Config |
 | :---------------- | :-------------------: | :------------------: | :-----: | :-----: | :------------: | :-----: | :-----------------------------------------------------: | :-----: |
-| w/o   |    4    |    1    | ResNet-r50-FPN      | [VisDrone](https://github.com/VisDrone/VisDrone-Dataset)  |   1x    |  23.3  | [Download Link](https://bj.bcebos.com/v1/paddledet/models/faster_rcnn_r50_fpn_1x_visdrone.pdparams ) | [Config](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.3/configs/sniper/faster_rcnn_r50_fpn_1x_sniper_coco.yml) |
-| w/ |    4    |    1    | ResNet-r50-FPN      | [VisDrone](https://github.com/VisDrone/VisDrone-Dataset)   |   1x    |  29.7  | [Download Link](https://bj.bcebos.com/v1/paddledet/models/faster_rcnn_r50_fpn_1x_sniper_visdrone.pdparams) | [Config](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.3/configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml) |
+| w/o   |    4    |    1    | ResNet-r50-FPN      | [VisDrone](https://github.com/VisDrone/VisDrone-Dataset)  |   1x    |  23.3  | [Download Link](https://bj.bcebos.com/v1/paddledet/models/faster_rcnn_r50_fpn_1x_visdrone.pdparams ) | [config](./faster_rcnn_r50_fpn_1x_visdrone.yml) |
+| w/ |    4    |    1    | ResNet-r50-FPN      | [VisDrone](https://github.com/VisDrone/VisDrone-Dataset)   |   1x    |  29.7  | [Download Link](https://bj.bcebos.com/v1/paddledet/models/faster_rcnn_r50_fpn_1x_sniper_visdrone.pdparams) | [config](./faster_rcnn_r50_fpn_1x_sniper_visdrone.yml) |
+
 
 ### Note
 > Here, we use VisDrone dataset, and to detect 9 objects including `person, bicycles, car, van, truck, tricyle, awning-tricyle, bus, motor`.
@@ -21,18 +22,18 @@ python tools/sniper_params_stats.py FasterRCNN annotations/instances_train2017.j
 ```
 b. optional: trian detector to get negative proposals.
 ```bash
-python -m paddle.distributed.launch --log_dir=./sniper/ --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml --save_proposals --proposals_path=./proposals.json &>sniper.log 2>&1 &
+python -m paddle.distributed.launch --log_dir=./sniper/ --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/sniper/faster_rcnn_r50_fpn_1x_sniper_visdrone.yml --save_proposals --proposals_path=./proposals.json &>sniper.log 2>&1 &
 ```
 c. train models
 ```bash
-python -m paddle.distributed.launch --log_dir=./sniper/ --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml --eval &>sniper.log 2>&1 &
+python -m paddle.distributed.launch --log_dir=./sniper/ --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/sniper/faster_rcnn_r50_fpn_1x_sniper_visdrone.yml --eval &>sniper.log 2>&1 &
 ```
 
 ### 2. Evaluation
 Evaluating SNIPER on custom dataset in single GPU with following commands:
 ```bash
 # use saved checkpoint in training
-CUDA_VISIBLE_DEVICES=0 python tools/eval.py -c configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml -o weights=output/faster_rcnn_r50_fpn_2x_sniper_coco/model_final
+CUDA_VISIBLE_DEVICES=0 python tools/eval.py -c configs/sniper/faster_rcnn_r50_fpn_1x_sniper_visdrone.yml -o weights=output/faster_rcnn_r50_fpn_1x_sniper_visdrone/model_final
 ```
 
 ### 3. Inference
@@ -40,10 +41,10 @@ Inference images in single GPU with following commands, use `--infer_img` to inf
 
 ```bash
 # inference single image
-CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml -o weights=output/faster_rcnn_r50_fpn_2x_sniper_coco/model_final --infer_img=demo/P0861__1.0__1154___824.png
+CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/sniper/faster_rcnn_r50_fpn_1x_sniper_visdrone.yml -o weights=output/faster_rcnn_r50_fpn_1x_sniper_visdrone/model_final --infer_img=demo/P0861__1.0__1154___824.png
 
 # inference all images in the directory
-CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml -o weights=output/faster_rcnn_r50_fpn_2x_sniper_coco/model_final --infer_dir=demo
+CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/sniper/faster_rcnn_r50_fpn_1x_sniper_visdrone.yml -o weights=output/faster_rcnn_r50_fpn_1x_sniper_visdrone/model_final --infer_dir=demo
 ```
 
 ## Citations

configs/sniper/README_cn.md

@@ -5,8 +5,8 @@
 ## 模型库
 | 有无sniper   | GPU个数    | 每张GPU图片个数 | 骨架网络             | 数据集     | 学习率策略 | Box AP |                           模型下载                          | 配置文件 |
 | :---------------- | :-------------------: | :------------------: | :-----: | :-----: | :------------: | :-----: | :-----------------------------------------------------: | :-----: |
-| w/o sniper   |    4    |    1    | ResNet-r50-FPN      | [VisDrone](https://github.com/VisDrone/VisDrone-Dataset)  |   1x    |  23.3  | [下载链接](https://bj.bcebos.com/v1/paddledet/models/faster_rcnn_r50_fpn_1x_visdrone.pdparams ) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.3/configs/sniper/faster_rcnn_r50_fpn_1x_sniper_coco.yml) |
-| w sniper |    4    |    1    | ResNet-r50-FPN      | [VisDrone](https://github.com/VisDrone/VisDrone-Dataset)    |   1x    |  29.7  | [下载链接](https://bj.bcebos.com/v1/paddledet/models/faster_rcnn_r50_fpn_1x_sniper_visdrone.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.3/configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml) |
+| w/o sniper   |    4    |    1    | ResNet-r50-FPN      | [VisDrone](https://github.com/VisDrone/VisDrone-Dataset)  |   1x    |  23.3  | [下载链接](https://bj.bcebos.com/v1/paddledet/models/faster_rcnn_r50_fpn_1x_visdrone.pdparams ) | [配置文件](./faster_rcnn_r50_fpn_1x_visdrone.yml) |
+| w sniper |    4    |    1    | ResNet-r50-FPN      | [VisDrone](https://github.com/VisDrone/VisDrone-Dataset)    |   1x    |  29.7  | [下载链接](https://bj.bcebos.com/v1/paddledet/models/faster_rcnn_r50_fpn_1x_sniper_visdrone.pdparams) | [配置文件](./faster_rcnn_r50_fpn_1x_sniper_visdrone.yml) |
 
 
 ### 注意
@@ -21,18 +21,18 @@ python tools/sniper_params_stats.py FasterRCNN annotations/instances_train2017.j
 ```
 b. 可选：训练检测器，生成负样本
 ```bash
-python -m paddle.distributed.launch --log_dir=./sniper/ --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml --save_proposals --proposals_path=./proposals.json &>sniper.log 2>&1 &
+python -m paddle.distributed.launch --log_dir=./sniper/ --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/sniper/faster_rcnn_r50_fpn_1x_sniper_visdrone.yml --save_proposals --proposals_path=./proposals.json &>sniper.log 2>&1 &
 ```
 c. 训练模型
 ```bash
-python -m paddle.distributed.launch --log_dir=./sniper/ --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml --eval &>sniper.log 2>&1 &
+python -m paddle.distributed.launch --log_dir=./sniper/ --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/sniper/faster_rcnn_r50_fpn_1x_sniper_visdrone.yml --eval &>sniper.log 2>&1 &
 ```
 
 ### 2. 评估
 使用单GPU通过如下命令一键式评估模型在COCO val2017数据集效果
 ```bash
 # 使用训练保存的checkpoint
-CUDA_VISIBLE_DEVICES=0 python tools/eval.py -c configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml -o weights=output/faster_rcnn_r50_fpn_2x_sniper_coco/model_final
+CUDA_VISIBLE_DEVICES=0 python tools/eval.py -c configs/sniper/faster_rcnn_r50_fpn_1x_sniper_visdrone.yml -o weights=output/faster_rcnn_r50_fpn_1x_sniper_visdrone/model_final
 ```
 
 ### 3. 推理
@@ -40,10 +40,10 @@ CUDA_VISIBLE_DEVICES=0 python tools/eval.py -c configs/sniper/faster_rcnn_r50_fp
 
 ```bash
 # 推理单张图像
-CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml -o weights=output/faster_rcnn_r50_fpn_2x_sniper_coco/model_final --infer_img=demo/P0861__1.0__1154___824.png
+CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/sniper/faster_rcnn_r50_fpn_1x_sniper_visdrone.yml -o weights=output/faster_rcnn_r50_fpn_1x_sniper_visdrone/model_final --infer_img=demo/P0861__1.0__1154___824.png
 
 # 推理目录下所有图像
-CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml -o weights=output/faster_rcnn_r50_fpn_2x_sniper_coco/model_final --infer_dir=demo
+CUDA_VISIBLE_DEVICES=0 python tools/infer.py -c configs/sniper/faster_rcnn_r50_fpn_1x_sniper_visdrone.yml -o weights=output/faster_rcnn_r50_fpn_1x_sniper_visdrone/model_final --infer_dir=demo
 ```
 
 ## Citations

configs/sniper/faster_rcnn_r50_fpn_1x_sniper_visdrone.yml

@@ -5,7 +5,27 @@ _BASE_: [
   '../faster_rcnn/_base_/optimizer_1x.yml',
   '_base_/faster_fpn_reader.yml',
 ]
-weights: output/faster_rcnn_r50_1x_visdrone_coco/model_final
+weights: output/faster_rcnn_r50_fpn_1x_sniper_visdrone/model_final
 find_unused_parameters: true
 
 
+
+metric: COCO
+num_classes: 9
+
+TrainDataset:
+  !COCODataSet
+    image_dir: train
+    anno_path: annotations/train.json
+    dataset_dir: dataset/VisDrone2019_coco
+    data_fields: ['image', 'gt_bbox', 'gt_class', 'is_crowd']
+
+EvalDataset:
+  !COCODataSet
+    image_dir: val
+    anno_path: annotations/val.json
+    dataset_dir: dataset/VisDrone2019_coco
+
+TestDataset:
+  !ImageFolder
+    anno_path: annotations/val.json

configs/faster_rcnn/faster_rcnn_r50_fpn_1x_visdrone.yml → configs/sniper/faster_rcnn_r50_fpn_1x_visdrone.yml

@@ -5,7 +5,7 @@ _BASE_: [
   '_base_/faster_rcnn_r50_fpn.yml',
   '_base_/faster_fpn_reader.yml',
 ]
-weights: output/faster_rcnn_r50_fpn_1x_coco_visdrone/model_final
+weights: output/faster_rcnn_r50_fpn_1x_visdrone/model_final
 
 
 metric: COCO

configs/sniper/faster_rcnn_r50_fpn_2x_sniper_coco.yml

-_BASE_: [
-  'faster_rcnn_r50_fpn_1x_sniper_coco.yml',
-]
-weights: output/faster_rcnn_r50_fpn_2x_sniper_coco/model_final
-
-epoch: 24
-LearningRate:
-  base_lr: 0.01
-  schedulers:
-  - !PiecewiseDecay
-    gamma: 0.1
-    milestones: [16, 22]
-  - !LinearWarmup
-    start_factor: 0.1
-    steps: 1000

configs/sniper/faster_rcnn_r50_vd_1x_coco_sniper.yml

-_BASE_: [
-  '../datasets/sniper_coco_detection.yml',
-  '../runtime.yml',
-  '../faster_rcnn/_base_/optimizer_1x.yml',
-  '../faster_rcnn/_base_/faster_rcnn_r50.yml',
-  '_base_/faster_reader.yml',
-]
-
-pretrain_weights: https://paddledet.bj.bcebos.com/models/faster_rcnn_r50_vd_1x_coco.pdparams
-weights: output/faster_rcnn_r50_vd_1x_coco/model_final
-
-ResNet:
-  # index 0 stands for res2
-  depth: 50
-  variant: d
-  norm_type: bn
-  freeze_at: 0
-  return_idx: [2]
-  num_stages: 3

configs/sniper/faster_rcnn_r50_vd_fpn_2x_sniper_coco.yml

-_BASE_: [
-  'faster_rcnn_r50_fpn_1x_sniper_coco.yml',
-]
-pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/ResNet50_vd_pretrained.pdparams
-weights: output/faster_rcnn_r50_vd_fpn_2x_sniper_coco/model_final
-
-ResNet:
-  # index 0 stands for res2
-  depth: 50
-  variant: d
-  norm_type: bn
-  freeze_at: 0
-  return_idx: [0,1,2,3]
-  num_stages: 4
-
-epoch: 24
-LearningRate:
-  base_lr: 0.01
-  schedulers:
-  - !PiecewiseDecay
-    gamma: 0.1
-    milestones: [16, 22]
-  - !LinearWarmup
-    start_factor: 0.1
-    steps: 1000

configs/sniper/ppyolo_r50vd_dcn_1x_sniper.yml

-_BASE_: [
-  '../datasets/sniper_coco_detection.yml',
-  '../runtime.yml',
-  '../ppyolo/_base_/ppyolo_r50vd_dcn.yml',
-  '../ppyolo/_base_/optimizer_1x.yml',
-  './_base_/ppyolo_reader.yml',
-]
-
-snapshot_epoch: 8
-use_ema: true
-weights: output/ppyolo_r50vd_dcn_1x_minicoco/model_final
-
-
-LearningRate:
-  base_lr: 0.005
-  schedulers:
-  - !PiecewiseDecay
-    gamma: 0.1
-    milestones:
-    - 153
-    - 173
-  - !LinearWarmup
-    start_factor: 0.
-    steps: 4000
-
-OptimizerBuilder:
-  optimizer:
-    momentum: 0.9
-    type: Momentum
-  regularizer:
-    factor: 0.0005
-    type: L2

configs/sniper/ppyolo_r50vd_dcn_1x_sniper_visdrone.yml

@@ -8,7 +8,7 @@ _BASE_: [
 
 snapshot_epoch: 8
 use_ema: true
-weights: output/ppyolo_r50vd_dcn_1x_visdrone/model_final
+weights: output/ppyolo_r50vd_dcn_1x_sniper_visdrone/model_final
 
 
 
@@ -31,3 +31,26 @@ OptimizerBuilder:
   regularizer:
     factor: 0.0005
     type: L2
+
+
+
+
+metric: COCO
+num_classes: 9
+
+TrainDataset:
+  !COCODataSet
+    image_dir: train
+    anno_path: annotations/train.json
+    dataset_dir: dataset/VisDrone2019_coco
+    data_fields: ['image', 'gt_bbox', 'gt_class', 'is_crowd']
+
+EvalDataset:
+  !COCODataSet
+    image_dir: val
+    anno_path: annotations/val.json
+    dataset_dir: dataset/VisDrone2019_coco
+
+TestDataset:
+  !ImageFolder
+    anno_path: annotations/val.json

configs/ppyolo/ppyolo_r50vd_dcn_1x_visdrone.yml → configs/sniper/ppyolo_r50vd_dcn_1x_visdrone.yml

@@ -8,7 +8,7 @@ _BASE_: [
 
 snapshot_epoch: 8
 use_ema: true
-weights: output/ppyolo_r50vd_dcn_1x_visdrone_coco/model_final
+weights: output/ppyolo_r50vd_dcn_1x_visdrone/model_final
 
 epoch: 192
 
@@ -51,4 +51,4 @@ EvalDataset:
 
 TestDataset:
   !ImageFolder
-    anno_path: annotations/val.json
\ No newline at end of file
+    anno_path: annotations/val.json