diff --git a/deploy/pphuman/README.md b/deploy/pphuman/README.md
index f5d324e7df4ec9f701323b0dc1db24b254333dbb..dbe3491c8934eac06acb20edff80ec3537f599f6 100644
--- a/deploy/pphuman/README.md
+++ b/deploy/pphuman/README.md
@@ -29,7 +29,8 @@ cd PaddleDetection
 pip install -r requirements.txt
 ```
 
-详细安装文档参考[文档](../../docs/tutorials/INSTALL_cn.md)
+1. 详细安装文档参考[文档](../../docs/tutorials/INSTALL_cn.md)
+2. 如果需要TensorRT推理加速（测速方式），请安装带`TensorRT版本Paddle`。您可以从[Paddle安装包](https://paddleinference.paddlepaddle.org.cn/v2.2/user_guides/download_lib.html#python)下载安装，或者按照[指导文档](https://www.paddlepaddle.org.cn/inference/master/optimize/paddle_trt.html)使用docker或自编译方式准备Paddle环境。
 
 ## 二、快速开始
 
@@ -93,21 +94,23 @@ ATTR:
 
 ```
 # 行人检测，指定配置文件路径和测试图片
-python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --image_file=test_image.jpg --device=gpu
+python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --image_file=test_image.jpg --device=gpu [--run_mode trt_fp16]
 
 # 行人跟踪，指定配置文件路径和测试视频
-python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu
+python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu [--run_mode trt_fp16]
 
 # 行人跟踪，指定配置文件路径，模型路径和测试视频
 # 命令行中指定的模型路径优先级高于配置文件
-python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu --model_dir det=ppyoloe/
+python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu --model_dir det=ppyoloe/ [--run_mode trt_fp16]
 
 # 行人属性识别，指定配置文件路径和测试视频
-python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu --enable_attr=True
+python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu --enable_attr=True [--run_mode trt_fp16]
 
 # 行为识别，指定配置文件路径和测试视频
-python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu --enable_action=True
+python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu --enable_action=True [--run_mode trt_fp16]
 
+# 行人跨境跟踪，指定配置文件路径和测试视频列表文件夹
+python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_dir=mtmct_dir/ --device=gpu [--run_mode trt_fp16]
 ```
 
 其他用法请参考[子任务文档](./docs)
diff --git a/deploy/pphuman/README_en.md b/deploy/pphuman/README_en.md
index b3531f0afd1b6816b31fda1313c5e542dab7c7d1..984a602999f81f48ed3a03a48b571313e811187b 100644
--- a/deploy/pphuman/README_en.md
+++ b/deploy/pphuman/README_en.md
@@ -28,7 +28,8 @@ cd PaddleDetection
 pip install -r requirements.txt
 ```
 
-For details of the installation, please refer to this [document](../../docs/tutorials/INSTALL.md)
+1. For details of the installation, please refer to this [document](../../docs/tutorials/INSTALL.md)
+2. Please install `Paddle-TensorRT` if your want speedup inference by TensorRT. You can download the whl package from [Paddle-whl-list](https://paddleinference.paddlepaddle.org.cn/v2.2/user_guides/download_lib.html#python), or prepare the envs by yourself follows the [Install-Guide](https://www.paddlepaddle.org.cn/inference/master/optimize/paddle_trt.html).
 
 ## II. Quick Start
 
@@ -91,23 +92,23 @@ ATTR:
 
 ```
 # Pedestrian detection. Specify the config file path and test images
-python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --image_file=test_image.jpg --device=gpu
+python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --image_file=test_image.jpg --device=gpu [--run_mode trt_fp16]
 
 # Pedestrian tracking. Specify the config file path and test videos
-python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu
+python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu [--run_mode trt_fp16]
 
 # Pedestrian tracking. Specify the config file path, the model path and test videos
 # The model path specified on the command line prioritizes over the config file
-python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu --model_dir det=ppyoloe/
+python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu --model_dir det=ppyoloe/ [--run_mode trt_fp16]
 
 # Attribute recognition. Specify the config file path and test videos
-python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu --enable_attr=True
+python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu --enable_attr=True [--run_mode trt_fp16]
 
 # Action Recognition. Specify the config file path and test videos
-python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu --enable_action=True
+python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_file=test_video.mp4 --device=gpu --enable_action=True [--run_mode trt_fp16]
 
-# Multi-Camera pedestrian tracking. Specify the config file path and test videos
-python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_dir=test_video_dir/ --device=gpu
+# Pedestrian Multi-Target Multi-Camera tracking. Specify the config file path and the directory of test videos
+python deploy/pphuman/pipeline.py --config deploy/pphuman/config/infer_cfg.yml --video_dir=mtmct_dir/ --device=gpu [--run_mode trt_fp16]
 
 ```
 
diff --git a/deploy/pphuman/mtmct.py b/deploy/pphuman/mtmct.py
index 5e0abbd9d0c7be69120cac04b3c5794d9bb9c436..5c7494edf1327955e045a8c82a2458bbda0a23de 100644
--- a/deploy/pphuman/mtmct.py
+++ b/deploy/pphuman/mtmct.py
@@ -113,8 +113,8 @@ def save_mtmct_vis_results(camera_results, captures, output_dir):
         cid = camera_ids[idx]
         basename = os.path.basename(video_file)
         video_out_name = "vis_" + basename
-        print("Start visualizing output video: {}".format(video_out_name))
         out_path = os.path.join(save_dir, video_out_name)
+        print("Start visualizing output video: {}".format(out_path))
 
         # Get Video info : resolution, fps, frame count
         width = int(capture.get(cv2.CAP_PROP_FRAME_WIDTH))
diff --git a/deploy/pphuman/pipeline.py b/deploy/pphuman/pipeline.py
index a9543357107cb48a6e7d4c3fbfdc88a8e11ff49e..8078ada54b38c0bd6d0f9ab22221bbbc13668836 100644
--- a/deploy/pphuman/pipeline.py
+++ b/deploy/pphuman/pipeline.py
@@ -109,8 +109,9 @@ class Pipeline(object):
         self.input = self._parse_input(image_file, image_dir, video_file,
                                        video_dir, camera_id)
         if self.multi_camera:
-            self.predictor = [
-                PipePredictor(
+            self.predictor = []
+            for name in self.input:
+                predictor_item = PipePredictor(
                     cfg,
                     is_video=True,
                     multi_camera=True,
@@ -123,8 +124,10 @@ class Pipeline(object):
                     trt_opt_shape=trt_opt_shape,
                     cpu_threads=cpu_threads,
                     enable_mkldnn=enable_mkldnn,
-                    output_dir=output_dir) for i in self.input
-            ]
+                    output_dir=output_dir)
+                predictor_item.set_file_name(name)
+                self.predictor.append(predictor_item)
+
         else:
             self.predictor = PipePredictor(
                 cfg,