[MOT]fix kitti cfgs and update kitti modelzoo (#4358)

configs/mot/kitticars/README_cn.md → configs/mot/vehicle/README_cn.md

@@ -15,7 +15,7 @@
 
 |    骨干网络      |  输入尺寸 |  MOTA   |   FPS   |  下载链接 | 配置文件 |
 | :--------------| :------- | :-----: | :-----: | :------: | :----: |
-| DLA-34         | 1088x608 |   53.9  |    -    |[下载链接](https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_1088x608_kitticars.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.2/configs/mot/kitticars/fairmot_dla34_30e_1088x608_kitticars.yml) |
+| DLA-34         | 1088x608 |   82.7  |    -    |[下载链接](https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_1088x608_kitti_vehicle.pdparams) | [配置文件](./fairmot_dla34_30e_1088x608_kitti_vehicle.yml) |
 
 **注意:**
  FairMOT使用2个GPU进行训练，每个GPU上batch size为6，训练30个epoch。
@@ -25,36 +25,36 @@
 ### 1. 训练
 使用2个GPU通过如下命令一键式启动训练
 ```bash
-python -m paddle.distributed.launch --log_dir=./fairmot_dla34_30e_1088x608_kitticars/ --gpus 0,1 tools/train.py -c configs/mot/kitticars/fairmot_dla34_30e_1088x608_kitticars.yml
+python -m paddle.distributed.launch --log_dir=./fairmot_dla34_30e_1088x608_kitti_vehicle/ --gpus 0,1 tools/train.py -c configs/mot/vehicle/fairmot_dla34_30e_1088x608_kitti_vehicle.yml
 ```
 
 ### 2. 评估
 使用单张GPU通过如下命令一键式启动评估
 ```bash
 # 使用PaddleDetection发布的权重
-CUDA_VISIBLE_DEVICES=0 python tools/eval_mot.py -c configs/mot/kitticars/fairmot_dla34_30e_1088x608_kitticars.yml -o weights=https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_1088x608_kitticars.pdparams
+CUDA_VISIBLE_DEVICES=0 python tools/eval_mot.py -c configs/mot/vehicle/fairmot_dla34_30e_1088x608_kitti_vehicle.yml -o weights=https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_1088x608_kitti_vehicle.pdparams
 
 # 使用训练保存的checkpoint
-CUDA_VISIBLE_DEVICES=0 python tools/eval_mot.py -c configs/mot/kitticars/fairmot_dla34_30e_1088x608_kitticars.yml -o weights=output/fairmot_dla34_30e_1088x608_kitticars/model_final.pdparams
+CUDA_VISIBLE_DEVICES=0 python tools/eval_mot.py -c configs/mot/vehicle/fairmot_dla34_30e_1088x608_kitti_vehicle.yml -o weights=output/fairmot_dla34_30e_1088x608_kitti_vehicle/model_final.pdparams
 ```
 
 ### 3. 预测
 使用单个GPU通过如下命令预测一个视频，并保存为视频
 ```bash
 # 预测一个视频
-CUDA_VISIBLE_DEVICES=0 python tools/infer_mot.py -c configs/mot/kitticars/fairmot_dla34_30e_1088x608_kitticars.yml -o weights=https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_1088x608_kitticars.pdparams --video_file={your video name}.mp4  --save_videos
+CUDA_VISIBLE_DEVICES=0 python tools/infer_mot.py -c configs/mot/vehicle/fairmot_dla34_30e_1088x608_kitti_vehicle.yml -o weights=https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_1088x608_kitti_vehicle.pdparams --video_file={your video name}.mp4  --save_videos
 ```
 **注意:**
  请先确保已经安装了[ffmpeg](https://ffmpeg.org/ffmpeg.html), Linux(Ubuntu)平台可以直接用以下命令安装：`apt-get update && apt-get install -y ffmpeg`。
 
 ### 4. 导出预测模型
 ```bash
-CUDA_VISIBLE_DEVICES=0 python tools/export_model.py -c configs/mot/kitticars/fairmot_dla34_30e_1088x608_kitticars.yml -o weights=https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_1088x608_kitticars.pdparams
+CUDA_VISIBLE_DEVICES=0 python tools/export_model.py -c configs/mot/kitti_vehicle/fairmot_dla34_30e_1088x608_kitti_vehicle.yml -o weights=https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_1088x608_kitti_vehicle.pdparams
 ```
 
 ### 5. 用导出的模型基于Python去预测
 ```bash
-python deploy/python/mot_jde_infer.py --model_dir=output_inference/fairmot_dla34_30e_1088x608_kitticars --video_file={your video name}.mp4 --device=GPU --save_mot_txts
+python deploy/python/mot_jde_infer.py --model_dir=output_inference/fairmot_dla34_30e_1088x608_kitti_vehicle --video_file={your video name}.mp4 --device=GPU --save_mot_txts
 ```
 **注意:**
  跟踪模型是对视频进行预测，不支持单张图的预测，默认保存跟踪结果可视化后的视频，可添加`--save_mot_txts`表示保存跟踪结果的txt文件，或`--save_images`表示保存跟踪结果可视化图片。

configs/mot/kitticars/fairmot_dla34_30e_1088x608_kitticars.yml → configs/mot/vehicle/fairmot_dla34_30e_1088x608_kitti_vehicle.yml

@@ -3,13 +3,13 @@ _BASE_: [
 ]
 
 metric: KITTI
-weights: output/fairmot_dla34_30e_1088x608_kitticars/model_final
+weights: output/fairmot_dla34_30e_1088x608_kitti_vehicle/model_final
 
 # for MOT training
 TrainDataset:
   !MOTDataSet
     dataset_dir: dataset/mot
-    image_lists: ['kitticars.train']
+    image_lists: ['kitti_vehicle.train']
     data_fields: ['image', 'gt_bbox', 'gt_class', 'gt_ide']
 
 # for MOT evaluation
@@ -17,7 +17,7 @@ TrainDataset:
 EvalMOTDataset:
   !MOTImageFolder
     dataset_dir: dataset/mot
-    data_root: kitticars/images/test
+    data_root: kitti_vehicle/images/train
     keep_ori_im: False # set True if save visualization images or video, or used in DeepSORT
 
 # for MOT video inference
@@ -25,3 +25,17 @@ TestMOTDataset:
   !MOTImageFolder
     dataset_dir: dataset/mot
     keep_ori_im: True # set True if save visualization images or video
+
+# model config
+FairMOT:
+  detector: CenterNet
+  reid: FairMOTEmbeddingHead
+  loss: FairMOTLoss
+  tracker: JDETracker
+
+JDETracker:
+  min_box_area: 200
+  vertical_ratio: 0 # no need to filter bboxes according to w/h
+  conf_thres: 0.4
+  tracked_thresh: 0.4
+  metric_type: cosine

deploy/python/mot_jde_infer.py

@@ -81,10 +81,14 @@ class JDE_Detector(Detector):
         assert batch_size == 1, "The JDE Detector only supports batch size=1 now"
         assert pred_config.tracker, "Tracking model should have tracker"
         tp = pred_config.tracker
+        min_box_area = tp['min_box_area'] if 'min_box_area' in tp else 200
+        vertical_ratio = tp['vertical_ratio'] if 'vertical_ratio' in tp else 1.6
         conf_thres = tp['conf_thres'] if 'conf_thres' in tp else 0.
         tracked_thresh = tp['tracked_thresh'] if 'tracked_thresh' in tp else 0.7
         metric_type = tp['metric_type'] if 'metric_type' in tp else 'euclidean'
         self.tracker = JDETracker(
+            min_box_area=min_box_area,
+            vertical_ratio=vertical_ratio,
             conf_thres=conf_thres,
             tracked_thresh=tracked_thresh,
             metric_type=metric_type)
@@ -100,11 +104,13 @@ class JDE_Detector(Detector):
             tid = t.track_id
             tscore = t.score
             if tscore < threshold: continue
-            vertical = tlwh[2] / tlwh[3] > 1.6
-            if tlwh[2] * tlwh[3] > self.tracker.min_box_area and not vertical:
-                online_tlwhs.append(tlwh)
-                online_ids.append(tid)
-                online_scores.append(tscore)
+            if tlwh[2] * tlwh[3] <= self.tracker.min_box_area: continue
+            if self.tracker.vertical_ratio > 0 and tlwh[2] / tlwh[
+                    3] > self.tracker.vertical_ratio:
+                continue
+            online_tlwhs.append(tlwh)
+            online_ids.append(tid)
+            online_scores.append(tscore)
         return online_tlwhs, online_scores, online_ids
 
     def predict(self, image_list, threshold=0.5, warmup=0, repeats=1):

deploy/python/tracker/jde_tracker.py

@@ -31,6 +31,9 @@ class JDETracker(object):
         det_thresh (float): threshold of detection score
         track_buffer (int): buffer for tracker
         min_box_area (int): min box area to filter out low quality boxes
+        vertical_ratio (float): w/h, the vertical ratio of the bbox to filter
+            bad results, set 1.6 default for pedestrian tracking. If set -1 
+            means no need to filter bboxes.
         tracked_thresh (float): linear assignment threshold of tracked 
             stracks and detections
         r_tracked_thresh (float): linear assignment threshold of 
@@ -47,6 +50,7 @@ class JDETracker(object):
                  det_thresh=0.3,
                  track_buffer=30,
                  min_box_area=200,
+                 vertical_ratio=1.6,
                  tracked_thresh=0.7,
                  r_tracked_thresh=0.5,
                  unconfirmed_thresh=0.7,
@@ -56,6 +60,8 @@ class JDETracker(object):
         self.det_thresh = det_thresh
         self.track_buffer = track_buffer
         self.min_box_area = min_box_area
+        self.vertical_ratio = vertical_ratio
+
         self.tracked_thresh = tracked_thresh
         self.r_tracked_thresh = r_tracked_thresh
         self.unconfirmed_thresh = unconfirmed_thresh

ppdet/engine/tracker.py

@@ -144,11 +144,13 @@ class Tracker(object):
                 tid = t.track_id
                 tscore = t.score
                 if tscore < draw_threshold: continue
-                vertical = tlwh[2] / tlwh[3] > 1.6
-                if tlwh[2] * tlwh[3] > tracker.min_box_area and not vertical:
-                    online_tlwhs.append(tlwh)
-                    online_ids.append(tid)
-                    online_scores.append(tscore)
+                if tlwh[2] * tlwh[3] <= tracker.min_box_area: continue
+                if tracker.vertical_ratio > 0 and tlwh[2] / tlwh[
+                        3] > tracker.vertical_ratio:
+                    continue
+                online_tlwhs.append(tlwh)
+                online_ids.append(tid)
+                online_scores.append(tscore)
             timer.toc()
 
             # save results

ppdet/metrics/mot_metrics.py

@@ -375,7 +375,7 @@ class KITTIEvaluation(object):
         # get number of sequences and
         # get number of frames per sequence from test mapping
         # (created while extracting the benchmark)
-        self.gt_path = os.path.join(gt_path, "label_02")
+        self.gt_path = os.path.join(gt_path, "../labels")
         self.n_frames = n_frames
         self.sequence_name = seqs
         self.n_sequences = n_sequences
@@ -1177,7 +1177,7 @@ class KITTIMOTMetric(Metric):
         assert data_type == 'kitti', "data_type should 'kitti'"
         self.result_root = result_root
         self.gt_path = data_root
-        gt_path = '{}/label_02/{}.txt'.format(data_root, seq)
+        gt_path = '{}/../labels/{}.txt'.format(data_root, seq)
         gt = open(gt_path, "r")
         max_frame = 0
         for line in gt:

ppdet/modeling/mot/tracker/jde_tracker.py

@@ -39,6 +39,9 @@ class JDETracker(object):
         det_thresh (float): threshold of detection score
         track_buffer (int): buffer for tracker
         min_box_area (int): min box area to filter out low quality boxes
+        vertical_ratio (float): w/h, the vertical ratio of the bbox to filter
+            bad results, set 1.6 default for pedestrian tracking. If set -1 
+            means no need to filter bboxes.
         tracked_thresh (float): linear assignment threshold of tracked 
             stracks and detections
         r_tracked_thresh (float): linear assignment threshold of 
@@ -55,6 +58,7 @@ class JDETracker(object):
                  det_thresh=0.3,
                  track_buffer=30,
                  min_box_area=200,
+                 vertical_ratio=1.6,
                  tracked_thresh=0.7,
                  r_tracked_thresh=0.5,
                  unconfirmed_thresh=0.7,
@@ -64,6 +68,8 @@ class JDETracker(object):
         self.det_thresh = det_thresh
         self.track_buffer = track_buffer
         self.min_box_area = min_box_area
+        self.vertical_ratio = vertical_ratio
+
         self.tracked_thresh = tracked_thresh
         self.r_tracked_thresh = r_tracked_thresh
         self.unconfirmed_thresh = unconfirmed_thresh