[MOT] update pptracking modelzoo (#4677)

* update pptracking modelzoo readme

* fix cfg, test=document_fix

configs/mot/fairmot/README.md

@@ -42,15 +42,16 @@ English | [简体中文](README_cn.md)
 | backbone       | input shape |  MOTA  |  IDF1  |  IDS  |   FP  |   FN   |   FPS   |  download | config |
 | :--------------| :------- | :----: | :----: | :----: | :----: | :----: | :------: | :----: |:-----: |
 | DLA-34         | 1088x608 |  75.9  |  74.7  |  1021   |  11425  |  31475 |    -     |[model](https://paddledet.bj.bcebos.com/models/mot/fairmot_enhance_dla34_60e_1088x608.pdparams) | [config](./fairmot_enhance_dla34_60e_1088x608.yml) |
+| HarDNet-85     | 1088x608 |  75.0  |  70.0  |  1050   |  11837  |  32774 |    -     |[model](https://paddledet.bj.bcebos.com/models/mot/fairmot_enhance_hardnet85_30e_1088x608.pdparams) | [config](./fairmot_enhance_hardnet85_30e_1088x608.yml) |
 
 ### Results on MOT-17 Test Set
 | backbone       | input shape |  MOTA  |  IDF1  |   IDS  |   FP   |   FN   |    FPS   |  download  | config |
 | :--------------| :------- | :----: | :----: | :----: | :----: | :----: | :------: | :----: |:-----: |
 | DLA-34         | 1088x608 |  75.3  |  74.2  |  3270  |  29112  | 106749 |    -     |[model](https://paddledet.bj.bcebos.com/models/mot/fairmot_enhance_dla34_60e_1088x608.pdparams) | [config](./fairmot_enhance_dla34_60e_1088x608.yml) |
+| HarDNet-85     | 1088x608 |  74.7  |  70.7  |  3210  |  29790  | 109914 |    -     |[model](https://paddledet.bj.bcebos.com/models/mot/fairmot_enhance_hardnet85_30e_1088x608.pdparams) | [config](./fairmot_enhance_hardnet85_30e_1088x608.yml) |
 
 **Notes:**
- FairMOT enhance DLA-34 used 8 GPUs for training and mini-batch size as 16 on each GPU，and trained for 60 epoches. The crowdhuman dataset is added to the train-set during training.
-
+ FairMOT enhance used 8 GPUs for training, and the crowdhuman dataset is added to the train-set during training. For FairMOT enhance DLA-34 the batch size is 16 on each GPU，and trained for 60 epoches. For FairMOT enhance HarDNet-85 the batch size is 10 on each GPU，and trained for 30 epoches. 
 
 ### FairMOT light model
 ### Results on MOT-16 Test Set

configs/mot/fairmot/README_cn.md

@@ -41,14 +41,16 @@
 |    骨干网络      |  输入尺寸 |  MOTA  |  IDF1  |  IDS  |   FP  |   FN   |   FPS   |  下载链接 | 配置文件 |
 | :--------------| :------- | :----: | :----: | :----: | :----: | :----: | :------: | :----: |:-----: |
 | DLA-34         | 1088x608 |  75.9  |  74.7  |  1021   |  11425  |  31475 |    -     |[下载链接](https://paddledet.bj.bcebos.com/models/mot/fairmot_enhance_dla34_60e_1088x608.pdparams) | [配置文件](./fairmot_enhance_dla34_60e_1088x608.yml) |
+| HarDNet-85     | 1088x608 |  75.0  |  70.0  |  1050   |  11837  |  32774 | -        |[下载链接](https://paddledet.bj.bcebos.com/models/mot/fairmot_enhance_hardnet85_30e_1088x608.pdparams) | [配置文件](./fairmot_enhance_hardnet85_30e_1088x608.yml) |
 
 ### 在MOT-17 Test Set上结果
 |    骨干网络      |  输入尺寸 |  MOTA  |  IDF1  |   IDS  |   FP   |   FN   |    FPS   |  下载链接  | 配置文件 |
 | :--------------| :------- | :----: | :----: | :----: | :----: | :----: | :------: | :----: |:-----: |
 | DLA-34         | 1088x608 |  75.3  |  74.2  |  3270  |  29112  | 106749 |    -     |[下载链接](https://paddledet.bj.bcebos.com/models/mot/fairmot_enhance_dla34_60e_1088x608.pdparams) | [配置文件](./fairmot_enhance_dla34_60e_1088x608.yml) |
+| HarDNet-85     | 1088x608 |  74.7  |  70.7  |  3210  |  29790  | 109914 |    -     |[下载链接](https://paddledet.bj.bcebos.com/models/mot/fairmot_enhance_hardnet85_30e_1088x608.pdparams) | [配置文件](./fairmot_enhance_hardnet85_30e_1088x608.yml) |
 
 **注意:**
- FairMOT enhance DLA-34使用8个GPU进行训练，每个GPU上batch size为16，训练60个epoch，并且训练集中加入了crowdhuman数据集一起参与训练。
+ FairMOT enhance模型均使用8个GPU进行训练，训练集中加入了crowdhuman数据集一起参与训练。DLA-34骨干网络的每个GPU上batch size为16，训练60个epoch。HarDNet-85骨干网络的每个GPU上batch size为10，训练30个epoch。
 
 
 ### FairMOT轻量级模型

configs/mot/fairmot/fairmot_enhance_hardnet85_30e_1088x608.yml

+_BASE_: [
+  '../../datasets/mot.yml',
+  '../../runtime.yml',
+  '_base_/optimizer_30e.yml',
+  '_base_/fairmot_hardnet85.yml',
+  '_base_/fairmot_reader_1088x608.yml',
+]
+norm_type: sync_bn
+use_ema: true
+ema_decay: 0.9998
+
+# add crowdhuman
+TrainDataset:
+  !MOTDataSet
+    dataset_dir: dataset/mot
+    image_lists: ['mot17.train', 'caltech.all', 'cuhksysu.train', 'prw.train', 'citypersons.train', 'eth.train', 'crowdhuman.train', 'crowdhuman.val']
+    data_fields: ['image', 'gt_bbox', 'gt_class', 'gt_ide']
+
+worker_num: 4
+TrainReader:
+  inputs_def:
+    image_shape: [3, 608, 1088]
+  sample_transforms:
+    - Decode: {}
+    - RGBReverse: {}
+    - AugmentHSV: {}
+    - LetterBoxResize: {target_size: [608, 1088]}
+    - MOTRandomAffine: {reject_outside: False}
+    - RandomFlip: {}
+    - BboxXYXY2XYWH: {}
+    - NormalizeBox: {}
+    - NormalizeImage: {mean: [0, 0, 0], std: [1, 1, 1]}
+    - RGBReverse: {}
+    - Permute: {}
+  batch_transforms:
+    - Gt2FairMOTTarget: {}
+  batch_size: 10
+  shuffle: True
+  drop_last: True
+  use_shared_memory: True
+
+epoch: 30
+LearningRate:
+  base_lr: 0.0001
+  schedulers:
+  - !PiecewiseDecay
+    gamma: 0.1
+    milestones: [20,]
+    use_warmup: False
+
+OptimizerBuilder:
+  optimizer:
+    type: Adam
+  regularizer: NULL
+
+weights: output/fairmot_enhance_hardnet85_30e_1088x608/model_final

configs/mot/headtracking21/README_cn.md

@@ -14,14 +14,18 @@
 |    骨干网络      |  输入尺寸 |  MOTA  |  IDF1  |  IDS  |   FP  |   FN   |   FPS   |  下载链接 | 配置文件 |
 | :--------------| :------- | :----: | :----: | :---: | :----: | :---: | :------: | :----: |:----: |
 | DLA-34         | 1088x608 |  64.7 |  69.0  |   8533  |  148817  |  234970  |     -   | [下载链接](https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_1088x608_headtracking21.pdparams) | [配置文件](./fairmot_dla34_30e_1088x608_headtracking21.yml) |
+| HRNetv2-W18    | 1088x608 |  57.2 |  58.4  |   30950 |  188260  |  256580  |     -   | [下载链接](https://paddledet.bj.bcebos.com/models/mot/fairmot_hrnetv2_w18_dlafpn_30e_1088x608_headtracking21.pdparams) | [配置文件](./fairmot_hrnetv2_w18_dlafpn_30e_1088x608_headtracking21.yml) |
+
 
 ### FairMOT在HT-21 Test Set上结果
 |    骨干网络      |  输入尺寸 |  MOTA  |  IDF1  |   IDS  |   FP   |   FN   |    FPS   |  下载链接  | 配置文件 |
 | :--------------| :------- | :----: | :----: | :----: | :----: | :----: |:-------: | :----: | :----: |
 | DLA-34         | 1088x608 |  60.8  |  62.8  |  12781   |  118109  |  198896 |    -     | [下载链接](https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_1088x608_headtracking21.pdparams) | [配置文件](./fairmot_dla34_30e_1088x608_headtracking21.yml) |
+| HRNetv2-W18    | 1088x608 |  41.2  |  47.1  |  48809   |  241683  |  204346 |    -     | [下载链接](https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_1088x608_headtracking21.pdparams) | [配置文件](./fairmot_dla34_30e_1088x608_headtracking21.yml) |
 
 **注意:**
  FairMOT DLA-34使用2个GPU进行训练，每个GPU上batch size为6，训练30个epoch。目前MOTA精度位于MOT官网[Head Tracking 21](https://motchallenge.net/results/Head_Tracking_21)榜单榜首。
+ FairMOT HRNetv2-W18使用4个GPU进行训练，每个GPU上batch size为8，训练30个epoch。
 
 ## 快速开始
 

configs/mot/mcfairmot/README.md

@@ -13,7 +13,7 @@ English | [简体中文](README_cn.md)
 MCFairMOT is the Multi-class extended version of [FairMOT](https://arxiv.org/abs/2004.01888).
 
 ## Model Zoo
-### MCFairMOT DLA-34 Results on VisDrone2019 Val Set
+### MCFairMOT Results on VisDrone2019 Val Set
 | backbone       | input shape | MOTA | IDF1 |  IDS    |   FPS    | download | config |
 | :--------------| :------- | :----: | :----: | :---:  | :------: | :----: |:----: |
 | DLA-34         | 1088x608 |  24.3  |  41.6  |  2314  |    -     |[model](https://paddledet.bj.bcebos.com/models/mot/mcfairmot_dla34_30e_1088x608_visdrone.pdparams) | [config](./mcfairmot_dla34_30e_1088x608_visdrone.yml) |
@@ -23,14 +23,15 @@ MCFairMOT is the Multi-class extended version of [FairMOT](https://arxiv.org/abs
 
 **Notes:**
  MOTA is the average MOTA of 10 catecories in the VisDrone2019 MOT dataset, and its value is also equal to the average MOTA of all the evaluated video sequences.
+ MCFairMOT used 4 GPUs for training 30 epoches. The batch size is 6 on each GPU for MCFairMOT DLA-34, and 4 for MCFairMOT HRNetV2-W18.
 
 
 ## Getting Start
 
 ### 1. Training
-Training MCFairMOT on 8 GPUs with following command
+Training MCFairMOT on 4 GPUs with following command
 ```bash
-python -m paddle.distributed.launch --log_dir=./mcfairmot_dla34_30e_1088x608_visdrone/ --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/mot/mcfairmot/mcfairmot_dla34_30e_1088x608_visdrone.yml
+python -m paddle.distributed.launch --log_dir=./mcfairmot_dla34_30e_1088x608_visdrone/ --gpus 0,1,2,3 tools/train.py -c configs/mot/mcfairmot/mcfairmot_dla34_30e_1088x608_visdrone.yml
 ```
 
 ### 2. Evaluation

configs/mot/mcfairmot/README_cn.md

@@ -14,7 +14,7 @@ MCFairMOT是[FairMOT](https://arxiv.org/abs/2004.01888)的多类别扩展版本
 
 ## 模型库
 
-### MCFairMOT DLA-34 在VisDrone2019 MOT val-set上结果
+### MCFairMOT 在VisDrone2019 MOT val-set上结果
 |    骨干网络      |  输入尺寸 |  MOTA  |  IDF1  |  IDS   |   FPS   |  下载链接 | 配置文件 |
 | :--------------| :------- | :----: | :----: | :---:  | :------: | :----: |:----: |
 | DLA-34         | 1088x608 |  24.3  |  41.6  |  2314  |    -     |[下载链接](https://paddledet.bj.bcebos.com/models/mot/mcfairmot_dla34_30e_1088x608_visdrone.pdparams) | [配置文件](./mcfairmot_dla34_30e_1088x608_visdrone.yml) |
@@ -24,13 +24,14 @@ MCFairMOT是[FairMOT](https://arxiv.org/abs/2004.01888)的多类别扩展版本
 
 **注意:**
  MOTA是VisDrone2019 MOT数据集10类目标的平均MOTA, 其值也等于所有评估的视频序列的平均MOTA。
+ MCFairMOT enhance模型均使用4个GPU进行训练，训练30个epoch。DLA-34骨干网络的每个GPU上batch size为6，HRNetV2-W18骨干网络的每个GPU上batch size为4。
 
 ## 快速开始
 
 ### 1. 训练
-使用8个GPU通过如下命令一键式启动训练
+使用4个GPU通过如下命令一键式启动训练
 ```bash
-python -m paddle.distributed.launch --log_dir=./mcfairmot_dla34_30e_1088x608_visdrone/ --gpus 0,1,2,3,4,5,6,7 tools/train.py -c configs/mot/mcfairmot/mcfairmot_dla34_30e_1088x608_visdrone.yml
+python -m paddle.distributed.launch --log_dir=./mcfairmot_dla34_30e_1088x608_visdrone/ --gpus 0,1,2,3 tools/train.py -c configs/mot/mcfairmot/mcfairmot_dla34_30e_1088x608_visdrone.yml
 ```
 
 ### 2. 评估

ppdet/modeling/necks/centernet_fpn.py

@@ -283,11 +283,10 @@ class TransitionUp(nn.Layer):
     def __init__(self, in_channels, out_channels):
         super().__init__()
 
-    def forward(self, x, skip, concat=True):
+    def forward(self, x, skip):
         w, h = skip.shape[2], skip.shape[3]
         out = F.interpolate(x, size=(w, h), mode="bilinear", align_corners=True)
-        if concat:
-            out = paddle.concat([out, skip], 1)
+        out = paddle.concat([out, skip], 1)
         return out
 
 
@@ -391,16 +390,17 @@ class CenterNetHarDNetFPN(nn.Layer):
 
         for i in range(3):
             skip_x = body_feats[3 - i]
-            x = self.transUpBlocks[i](x, skip_x, (i < self.skip_lv))
-            x = self.conv1x1_up[i](x)
+            x_up = self.transUpBlocks[i](x, skip_x)
+            x_ch = self.conv1x1_up[i](x_up)
             if self.SC[i] > 0:
-                end = x.shape[1]
-                x_sc.append(x[:, end - self.SC[i]:, :, :])
-                x = x[:, :end - self.SC[i], :, :]
-            x2 = self.avg9x9(x)
-            x3 = x / (x.sum((2, 3), keepdim=True) + 0.1)
-            x = paddle.concat([x, x2, x3], 1)
-            x = self.denseBlocksUp[i](x)
+                end = x_ch.shape[1]
+                new_st = end - self.SC[i]
+                x_sc.append(x_ch[:, new_st:, :, :])
+                x_ch = x_ch[:, :new_st, :, :]
+            x2 = self.avg9x9(x_ch)
+            x3 = x_ch / (x_ch.sum((2, 3), keepdim=True) + 0.1)
+            x_new = paddle.concat([x_ch, x2, x3], 1)
+            x = self.denseBlocksUp[i](x_new)
 
         scs = [x]
         for i in range(3):