Add python deployment for squeezesegv3

deploy/squeezesegv3/python/infer.py

@@ -0,0 +1,164 @@
+# Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import argparse
+
+import cv2
+import numpy as np
+import paddle
+from paddle.inference import Config, create_predictor
+
+from paddle3d import transforms as T
+from paddle3d.sample import Sample
+from paddle3d.transforms.normalize import NormalizeRangeImage
+from paddle3d.transforms.reader import LoadSemanticKITTIRange
+
+
+def parse_args():
+ parser = argparse.ArgumentParser()
+ parser.add_argument(
+ "--model_file",
+ type=str,
+ help="Model filename, Specify this when your model is a combined model.",
+ required=True)
+ parser.add_argument(
+ "--params_file",
+ type=str,
+ help=
+ "Parameter filename, Specify this when your model is a combined model.",
+ required=True)
+ parser.add_argument(
+ '--lidar_file', type=str, help='The lidar path.', required=True)
+ parser.add_argument(
+ '--img_mean',
+ type=str,
+ help='The mean value of range-view image.',
+ required=True)
+ parser.add_argument(
+ '--img_std',
+ type=str,
+ help='The variance value of range-view image.',
+ required=True)
+ parser.add_argument("--gpu_id", type=int, default=0, help="GPU card id.")
+ parser.add_argument(
+ "--use_trt",
+ type=int,
+ default=0,
+ help="Whether to use tensorrt to accelerate when using gpu.")
+ parser.add_argument(
+ "--trt_precision",
+ type=int,
+ default=0,
+ help="Precision type of tensorrt, 0: kFloat32, 1: kHalf.")
+ parser.add_argument(
+ "--trt_use_static",
+ type=int,
+ default=0,
+ help="Whether to load the tensorrt graph optimization from a disk path."
+ )
+ parser.add_argument(
+ "--trt_static_dir",
+ type=str,
+ help="Path of a tensorrt graph optimization directory.")
+
+ return parser.parse_args()
+
+
+def preprocess(file_path, img_mean, img_std):
+ if isinstance(img_mean, str):
+ img_mean = eval(img_mean)
+ if isinstance(img_std, str):
+ img_std = eval(img_std)
+
+ sample = Sample(path=file_path, modality="lidar")
+
+ transforms = T.Compose([
+ LoadSemanticKITTIRange(project_label=False),
+ NormalizeRangeImage(mean=img_mean, std=img_std)
+ ])
+
+ sample = transforms(sample)
+
+ if "proj_mask" in sample.meta:
+ sample.data *= sample.meta.pop("proj_mask")
+ return np.expand_dims(sample.data,
+ 0), sample.meta.proj_x, sample.meta.proj_y
+
+
+def init_predictor(model_file,
+ params_file,
+ gpu_id=0,
+ use_trt=False,
+ trt_precision=0,
+ trt_use_static=False,
+ trt_static_dir=None):
+ config = Config(model_file, params_file)
+ config.enable_memory_optim()
+ config.enable_use_gpu(1000, gpu_id)
+ if use_trt:
+ precision_mode = paddle.inference.PrecisionType.Float32
+ if trt_precision == 1:
+ precision_mode = paddle.inference.PrecisionType.Half
+ config.enable_tensorrt_engine(
+ workspace_size=1 << 20,
+ max_batch_size=1,
+ min_subgraph_size=3,
+ precision_mode=precision_mode,
+ use_static=trt_use_static,
+ use_calib_mode=False)
+ if trt_use_static:
+ config.set_optim_cache_dir(trt_static_dir)
+
+ predictor = create_predictor(config)
+ return predictor
+
+
+def run(predictor, points):
+ # copy img data to input tensor
+ input_names = predictor.get_input_names()
+ input_tensor = predictor.get_input_handle(input_names[0])
+ input_tensor.reshape(points.shape)
+ input_tensor.copy_from_cpu(points.copy())
+
+ # do the inference
+ predictor.run()
+
+ results = []
+ # get out data from output tensor
+ output_names = predictor.get_output_names()
+ output_tensor = predictor.get_output_handle(output_names[0])
+ pred_label = output_tensor.copy_to_cpu()
+
+ return pred_label[0]
+
+
+def postprocess(pred_img_label, proj_x, proj_y):
+ return pred_img_label[proj_y, proj_x]
+
+
+def main(args):
+ predictor = init_predictor(args.model_file, args.params_file, args.gpu_id,
+ args.use_trt, args.trt_precision,
+ args.trt_use_static, args.trt_static_dir)
+ range_img, proj_x, proj_y = preprocess(args.lidar_file, args.img_mean,
+ args.img_std)
+ pred_img_label = run(predictor, range_img)
+ pred_point_label = postprocess(pred_img_label, proj_x, proj_y)
+ return pred_point_label
+
+
+if __name__ == '__main__':
+ args = parse_args()
+
+ main(args)

docs/models/squeezesegv3/README.md

@@ -10,6 +10,7 @@
  * [训练](#h3-id52h3)
  * [评估](#h3-id53h3)
  * [模型导出](#h3-id54h3)
+ * [模型部署](#h3-id55h3)
 
 ## <h2 id="1">引用</h2>
 
@@ -125,3 +126,42 @@ python tools/export.py \
 | model | 待导出模型参数`model.pdparams`路径 | 是 | - |
 | input_shape | 指定模型的输入尺寸，支持`N, C, H, W`或`H, W`格式 | 是 | - |
 | save_dir | 保存导出模型的路径，`save_dir`下将会生成三个文件：`squeezesegv3.pdiparams `、`squeezesegv3.pdiparams.info`和`squeezesegv3.pdmodel` | 否 | `deploy` |
+
+
+
+### <h3 id="55">模型部署</h3>
+
+#### C++部署
+
+Coming soon...
+
+#### Python部署
+
+命令参数说明如下：
+
+| 参数 | 说明 |
+| -- | -- |
+| model_file | 导出模型的结构文件`squeezesegv3.pdmodel`所在路径 |
+| params_file | 导出模型的参数文件`squeezesegv3.pdiparams`所在路径 |
+| lidar_file | 待预测的点云文件所在路径 |
+| img_mean | 点云投影到range-view后所成图像的均值，例如为`12.12,10.88,0.23,-1.04,0.21` |
+| img_std | 点云投影到range-view后所成图像的方差，例如为`12.32,11.47,6.91,0.86,0.16` |
+| use_trt | 是否使用TensorRT进行加速，默认0|
+| trt_precision | 当use_trt设置为1时，模型精度可设置0或1，0表示fp32, 1表示fp16。默认0 |
+| trt_use_static | 当trt_use_static设置为1时，**在首次运行程序的时候会将TensorRT的优化信息进行序列化到磁盘上，下次运行时直接加载优化的序列化信息而不需要重新生成**。默认0 |
+| trt_static_dir | 当trt_use_static设置为1时，保存优化信息的路径 |
+
+
+运行以下命令，执行预测：
+
+```
+python infer.py --model_file /path/to/squeezesegv3.pdmodel --params_file /path/to/squeezesegv3.pdiparams --lidar_file /path/to/lidar.pcd.bin --img_mean 12.12,10.88,0.23,-1.04,0.21 --img_std 12.32,11.47,6.91,0.86,0.16
+```
+
+如果要开启TensorRT的话，请卸载掉原有的`paddlepaddel_gpu`，至[Paddle官网](https://paddleinference.paddlepaddle.org.cn/user_guides/download_lib.html#python)下载与TensorRT连编的预编译Paddle Inferece安装包，选择符合本地环境CUDA/cuDNN/TensorRT版本的安装包完成安装即可。
+
+运行以下命令，开启TensorRT加速模型预测：
+
+```
+python infer.py --model_file /path/to/squeezesegv3.pdmodel --params_file /path/to/squeezesegv3.pdiparams --lidar_file /path/to/lidar.pcd.bin --img_mean 12.12,10.88,0.23,-1.04,0.21 --img_std 12.32,11.47,6.91,0.86,0.16 --use_trt 1
+```