style(pre-commit): autofix

perception/autoware_bytetrack/package.xml

@@ -15,10 +15,10 @@
   <buildtool_export_depend>cudnn_cmake_module</buildtool_export_depend>
   <buildtool_export_depend>tensorrt_cmake_module</buildtool_export_depend>
 
+  <depend>autoware_cuda_utils</depend>
   <depend>autoware_kalman_filter</depend>
   <depend>autoware_perception_msgs</depend>
   <depend>autoware_tensorrt_common</depend>
-  <depend>autoware_cuda_utils</depend>
   <depend>cv_bridge</depend>
   <depend>eigen</depend>
   <depend>image_transport</depend>

perception/autoware_lidar_apollo_instance_segmentation/include/autoware/lidar_apollo_instance_segmentation/detector.hpp

@@ -19,10 +19,10 @@
 #include "autoware/lidar_apollo_instance_segmentation/feature_generator.hpp"
 #include "autoware/lidar_apollo_instance_segmentation/node.hpp"
 
-#include <autoware/tensorrt_common/tensorrt_common.hpp>
-#include <autoware/universe_utils/transform/transforms.hpp>
 #include <autoware/cuda_utils/cuda_unique_ptr.hpp>
 #include <autoware/cuda_utils/stream_unique_ptr.hpp>
+#include <autoware/tensorrt_common/tensorrt_common.hpp>
+#include <autoware/universe_utils/transform/transforms.hpp>
 #include <tf2_eigen/tf2_eigen.hpp>
 
 #include <tf2_ros/buffer_interface.h>

perception/autoware_lidar_apollo_instance_segmentation/package.xml

@@ -13,10 +13,10 @@
 
   <buildtool_depend>ament_cmake</buildtool_depend>
 
+  <depend>autoware_cuda_utils</depend>
   <depend>autoware_perception_msgs</depend>
   <depend>autoware_tensorrt_common</depend>
   <depend>autoware_universe_utils</depend>
-  <depend>autoware_cuda_utils</depend>
   <depend>libpcl-all-dev</depend>
   <depend>pcl_conversions</depend>
   <depend>rclcpp</depend>

perception/autoware_shape_estimation/include/autoware/shape_estimation/tensorrt_shape_estimator.hpp

@@ -15,10 +15,10 @@
 #ifndef AUTOWARE__SHAPE_ESTIMATION__TENSORRT_SHAPE_ESTIMATOR_HPP_
 #define AUTOWARE__SHAPE_ESTIMATION__TENSORRT_SHAPE_ESTIMATOR_HPP_
 
-#include <autoware/tensorrt_common/tensorrt_common.hpp>
 #include <autoware/cuda_utils/cuda_check_error.hpp>
 #include <autoware/cuda_utils/cuda_unique_ptr.hpp>
 #include <autoware/cuda_utils/stream_unique_ptr.hpp>
+#include <autoware/tensorrt_common/tensorrt_common.hpp>
 #include <pcl_ros/transforms.hpp>
 
 #include <autoware_perception_msgs/msg/detected_objects.hpp>

perception/autoware_shape_estimation/lib/tensorrt_shape_estimator.cpp

@@ -47,7 +47,8 @@ TrtShapeEstimator::TrtShapeEstimator(
   const auto one_hot_input_size = std::accumulate(
     one_hot_input_dims.d + 1, one_hot_input_dims.d + one_hot_input_dims.nbDims, 1,
     std::multiplies<int>());
-  input_one_hot_d_ = autoware::cuda_utils::make_unique<float[]>(one_hot_input_size * batch_config[2]);
+  input_one_hot_d_ =
+    autoware::cuda_utils::make_unique<float[]>(one_hot_input_size * batch_config[2]);
 
   const auto stage1_center_out_dims = trt_common_->getBindingDimensions(2);
   out_s1center_elem_num_ = std::accumulate(

perception/autoware_shape_estimation/package.xml

@@ -17,10 +17,10 @@
   <buildtool_export_depend>cudnn_cmake_module</buildtool_export_depend>
   <buildtool_export_depend>tensorrt_cmake_module</buildtool_export_depend>
 
+  <depend>autoware_cuda_utils</depend>
   <depend>autoware_perception_msgs</depend>
   <depend>autoware_tensorrt_common</depend>
   <depend>autoware_universe_utils</depend>
-  <depend>autoware_cuda_utils</depend>
   <depend>eigen</depend>
   <depend>libopencv-dev</depend>
   <depend>libpcl-all-dev</depend>

perception/autoware_tensorrt_classifier/include/autoware/tensorrt_classifier/tensorrt_classifier.hpp

@@ -15,10 +15,10 @@
 #ifndef AUTOWARE__TENSORRT_CLASSIFIER__TENSORRT_CLASSIFIER_HPP_
 #define AUTOWARE__TENSORRT_CLASSIFIER__TENSORRT_CLASSIFIER_HPP_
 
-#include <autoware/tensorrt_common/tensorrt_common.hpp>
 #include <autoware/cuda_utils/cuda_check_error.hpp>
 #include <autoware/cuda_utils/cuda_unique_ptr.hpp>
 #include <autoware/cuda_utils/stream_unique_ptr.hpp>
+#include <autoware/tensorrt_common/tensorrt_common.hpp>
 #include <opencv2/opencv.hpp>
 
 #include <memory>

perception/autoware_tensorrt_classifier/package.xml

@@ -19,8 +19,8 @@
 
   <build_depend>autoware_cmake</build_depend>
 
-  <depend>autoware_tensorrt_common</depend>
   <depend>autoware_cuda_utils</depend>
+  <depend>autoware_tensorrt_common</depend>
   <depend>libopencv-dev</depend>
   <depend>rclcpp</depend>
 

perception/autoware_tensorrt_classifier/src/tensorrt_classifier.cpp

@@ -176,7 +176,8 @@ TrtClassifier::TrtClassifier(
   out_elem_num_ = out_elem_num_ * batch_config[2];
   out_elem_num_per_batch_ = static_cast<int>(out_elem_num_ / batch_config[2]);
   out_prob_d_ = autoware::cuda_utils::make_unique<float[]>(out_elem_num_);
-  out_prob_h_ = autoware::cuda_utils::make_unique_host<float[]>(out_elem_num_, cudaHostAllocPortable);
+  out_prob_h_ =
+    autoware::cuda_utils::make_unique_host<float[]>(out_elem_num_, cudaHostAllocPortable);
 
   if (cuda) {
     m_cuda = true;

perception/autoware_tensorrt_yolox/include/autoware/tensorrt_yolox/tensorrt_yolox.hpp

@@ -15,10 +15,10 @@
 #ifndef AUTOWARE__TENSORRT_YOLOX__TENSORRT_YOLOX_HPP_
 #define AUTOWARE__TENSORRT_YOLOX__TENSORRT_YOLOX_HPP_
 
-#include <autoware/tensorrt_common/tensorrt_common.hpp>
-#include <autoware/tensorrt_yolox/preprocess.hpp>
 #include <autoware/cuda_utils/cuda_unique_ptr.hpp>
 #include <autoware/cuda_utils/stream_unique_ptr.hpp>
+#include <autoware/tensorrt_common/tensorrt_common.hpp>
+#include <autoware/tensorrt_yolox/preprocess.hpp>
 #include <opencv2/opencv.hpp>
 
 #include <memory>

perception/autoware_tensorrt_yolox/package.xml

@@ -19,10 +19,10 @@
   <buildtool_export_depend>cudnn_cmake_module</buildtool_export_depend>
   <buildtool_export_depend>tensorrt_cmake_module</buildtool_export_depend>
 
+  <depend>autoware_cuda_utils</depend>
   <depend>autoware_object_recognition_utils</depend>
   <depend>autoware_perception_msgs</depend>
   <depend>autoware_tensorrt_common</depend>
-  <depend>autoware_cuda_utils</depend>
   <depend>cv_bridge</depend>
   <depend>image_transport</depend>
   <depend>libopencv-dev</depend>

perception/autoware_tensorrt_yolox/src/tensorrt_yolox.cpp

@@ -283,7 +283,8 @@ TrtYoloX::TrtYoloX(
     out_elem_num_ = out_elem_num_ * batch_config[2];
     out_elem_num_per_batch_ = static_cast<int>(out_elem_num_ / batch_config[2]);
     out_prob_d_ = autoware::cuda_utils::make_unique<float[]>(out_elem_num_);
-    out_prob_h_ = autoware::cuda_utils::make_unique_host<float[]>(out_elem_num_, cudaHostAllocPortable);
+    out_prob_h_ =
+      autoware::cuda_utils::make_unique_host<float[]>(out_elem_num_, cudaHostAllocPortable);
     int w = input_dims.d[3];
     int h = input_dims.d[2];
     int sum_tensors = (w / 8) * (h / 8) + (w / 16) * (h / 16) + (w / 32) * (h / 32);
@@ -300,9 +301,11 @@ TrtYoloX::TrtYoloX(
     max_detections_ = out_scores_dims.d[1];
     input_d_ = autoware::cuda_utils::make_unique<float[]>(batch_config[2] * input_size);
     out_num_detections_d_ = autoware::cuda_utils::make_unique<int32_t[]>(batch_config[2]);
-    out_boxes_d_ = autoware::cuda_utils::make_unique<float[]>(batch_config[2] * max_detections_ * 4);
+    out_boxes_d_ =
+      autoware::cuda_utils::make_unique<float[]>(batch_config[2] * max_detections_ * 4);
     out_scores_d_ = autoware::cuda_utils::make_unique<float[]>(batch_config[2] * max_detections_);
-    out_classes_d_ = autoware::cuda_utils::make_unique<int32_t[]>(batch_config[2] * max_detections_);
+    out_classes_d_ =
+      autoware::cuda_utils::make_unique<int32_t[]>(batch_config[2] * max_detections_);
   }
   if (multitask_) {
     // Allocate buffer for segmentation
@@ -317,9 +320,10 @@ TrtYoloX::TrtYoloX(
     }
     segmentation_out_elem_num_per_batch_ =
       static_cast<int>(segmentation_out_elem_num_ / batch_config[2]);
-    segmentation_out_prob_d_ = autoware::cuda_utils::make_unique<float[]>(segmentation_out_elem_num_);
-    segmentation_out_prob_h_ =
-      autoware::cuda_utils::make_unique_host<float[]>(segmentation_out_elem_num_, cudaHostAllocPortable);
+    segmentation_out_prob_d_ =
+      autoware::cuda_utils::make_unique<float[]>(segmentation_out_elem_num_);
+    segmentation_out_prob_h_ = autoware::cuda_utils::make_unique_host<float[]>(
+      segmentation_out_elem_num_, cudaHostAllocPortable);
   }
   if (use_gpu_preprocess) {
     use_gpu_preprocess_ = true;
@@ -396,7 +400,8 @@ void TrtYoloX::initPreprocessBuffer(int width, int height)
       }
       image_buf_h_ = autoware::cuda_utils::make_unique_host<unsigned char[]>(
         width * height * 3 * batch_size_, cudaHostAllocWriteCombined);
-      image_buf_d_ = autoware::cuda_utils::make_unique<unsigned char[]>(width * height * 3 * batch_size_);
+      image_buf_d_ =
+        autoware::cuda_utils::make_unique<unsigned char[]>(width * height * 3 * batch_size_);
     }
     if (multitask_) {
       size_t argmax_out_elem_num = 0;
@@ -414,8 +419,8 @@ void TrtYoloX::initPreprocessBuffer(int width, int height)
         size_t out_elem_num = out_w * out_h * batch_size_;
         argmax_out_elem_num += out_elem_num;
       }
-      argmax_buf_h_ =
-        autoware::cuda_utils::make_unique_host<unsigned char[]>(argmax_out_elem_num, cudaHostAllocPortable);
+      argmax_buf_h_ = autoware::cuda_utils::make_unique_host<unsigned char[]>(
+        argmax_out_elem_num, cudaHostAllocPortable);
       argmax_buf_d_ = autoware::cuda_utils::make_unique<unsigned char[]>(argmax_out_elem_num);
     }
   }
@@ -470,8 +475,8 @@ void TrtYoloX::preprocessGpu(const std::vector<cv::Mat> & images)
       scales_.emplace_back(scale);
       image_buf_h_ = autoware::cuda_utils::make_unique_host<unsigned char[]>(
         image.cols * image.rows * 3 * batch_size, cudaHostAllocWriteCombined);
-      image_buf_d_ =
-        autoware::cuda_utils::make_unique<unsigned char[]>(image.cols * image.rows * 3 * batch_size);
+      image_buf_d_ = autoware::cuda_utils::make_unique<unsigned char[]>(
+        image.cols * image.rows * 3 * batch_size);
     }
     int index = b * image.cols * image.rows * 3;
     // Copy into pinned memory
@@ -496,8 +501,8 @@ void TrtYoloX::preprocessGpu(const std::vector<cv::Mat> & images)
 
   if (multitask_) {
     if (!argmax_buf_h_) {
-      argmax_buf_h_ =
-        autoware::cuda_utils::make_unique_host<unsigned char[]>(argmax_out_elem_num, cudaHostAllocPortable);
+      argmax_buf_h_ = autoware::cuda_utils::make_unique_host<unsigned char[]>(
+        argmax_out_elem_num, cudaHostAllocPortable);
     }
     if (!argmax_buf_d_) {
       argmax_buf_d_ = autoware::cuda_utils::make_unique<unsigned char[]>(argmax_out_elem_num);
@@ -618,8 +623,8 @@ void TrtYoloX::preprocessWithRoiGpu(
     if (!image_buf_h_) {
       image_buf_h_ = autoware::cuda_utils::make_unique_host<unsigned char[]>(
         image.cols * image.rows * 3 * batch_size, cudaHostAllocWriteCombined);
-      image_buf_d_ =
-        autoware::cuda_utils::make_unique<unsigned char[]>(image.cols * image.rows * 3 * batch_size);
+      image_buf_d_ = autoware::cuda_utils::make_unique<unsigned char[]>(
+        image.cols * image.rows * 3 * batch_size);
     }
     int index = b * image.cols * image.rows * 3;
     // Copy into pinned memory
@@ -735,7 +740,8 @@ void TrtYoloX::multiScalePreprocessGpu(const cv::Mat & image, const std::vector<
   if (!image_buf_h_) {
     image_buf_h_ = autoware::cuda_utils::make_unique_host<unsigned char[]>(
       image.cols * image.rows * 3 * 1, cudaHostAllocWriteCombined);
-    image_buf_d_ = autoware::cuda_utils::make_unique<unsigned char[]>(image.cols * image.rows * 3 * 1);
+    image_buf_d_ =
+      autoware::cuda_utils::make_unique<unsigned char[]>(image.cols * image.rows * 3 * 1);
   }
   int index = 0 * image.cols * image.rows * 3;
   // Copy into pinned memory

perception/autoware_traffic_light_classifier/package.xml

@@ -14,9 +14,9 @@
 
   <build_depend>autoware_cmake</build_depend>
 
+  <depend>autoware_cuda_utils</depend>
   <depend>autoware_tensorrt_classifier</depend>
   <depend>autoware_tensorrt_common</depend>
-  <depend>autoware_cuda_utils</depend>
   <depend>cv_bridge</depend>
   <depend>image_transport</depend>
   <depend>libboost-filesystem-dev</depend>

perception/autoware_traffic_light_classifier/src/classifier/cnn_classifier.hpp

@@ -17,10 +17,10 @@
 
 #include "classifier_interface.hpp"
 
-#include <autoware/tensorrt_classifier/tensorrt_classifier.hpp>
-#include <autoware/tensorrt_common/tensorrt_common.hpp>
 #include <autoware/cuda_utils/cuda_unique_ptr.hpp>
 #include <autoware/cuda_utils/stream_unique_ptr.hpp>
+#include <autoware/tensorrt_classifier/tensorrt_classifier.hpp>
+#include <autoware/tensorrt_common/tensorrt_common.hpp>
 #include <image_transport/image_transport.hpp>
 #include <opencv2/core/core.hpp>
 #include <opencv2/highgui/highgui.hpp>