fix: refactor files

Signed-off-by: Kaan Çolak <kaancolak95@gmail.com>

Author: kaancolak

projects/AutowareCenterPoint/centerpoint/pillar_encoder_autoware.py

@@ -1,8 +1,6 @@
 from typing import Optional, Tuple
 
 import torch
-from mmcv.cnn import build_norm_layer
-from mmcv.ops import DynamicScatter
 from torch import Tensor, nn
 
 from mmdet3d.models.voxel_encoders.utils import (PFNLayer,

projects/AutowareCenterPoint/centerpoint_onnx_converter.py

@@ -1,6 +1,6 @@
 import argparse
 import os
-from typing import Any, Dict, List, Optional, Tuple
+from typing import Dict, List, Optional, Tuple
 
 import torch
 from mmengine import Config
@@ -17,8 +17,7 @@
 
 def parse_args():
     parser = argparse.ArgumentParser(
-        description=
-        'Create autoware compitable onnx file from torch checkpoint ')
+        'Create autoware compatible onnx file from torch checkpoint ')
     parser.add_argument('--cfg', help='train config file path')
     parser.add_argument('--ckpt', help='checkpoint weeight')
     parser.add_argument('--work-dir', help='the dir to save onnx files')
@@ -119,9 +118,8 @@ def save_onnx(self) -> None:
             verbose=False,
             opset_version=11,
         )
-        print(
-            f'Saved pts_backbone_neck_head onnx model: {pth_onnx_backbone_neck_head}'
-        )
+        print(f'Saved pts_backbone_neck_head onnx model:'
+              f' {pth_onnx_backbone_neck_head}')
 
 
 @MODELS.register_module()
@@ -280,7 +278,8 @@ def forward(self, x: torch.Tensor) -> Tuple[List[Dict[str, torch.Tensor]]]:
             x (torch.Tensor): (B, C, H, W)
         Returns:
             tuple[list[dict[str, any]]]:
-                (num_classes x [num_detect x {'reg', 'height', 'dim', 'rot', 'vel', 'heatmap'}])
+                (num_classes x [num_detect x
+                {'reg', 'height', 'dim', 'rot', 'vel', 'heatmap'}])
         """
         x = self.backbone(x)
         if self.neck is not None:

projects/AutowareCenterPoint/configs/centerpoint_custom.py

@@ -845,4 +845,5 @@
     0.2,
     8,
 ]
-work_dir = './work_dirs/centerpoint_pillar02_second_secfpn_head-circlenms_8xb4-cyclic-20e_nus-3d-CUSTOM'
+work_dir = ('./work_dirs/centerpoint_pillar02_second_'
+            'secfpn_head-circlenms_8xb4-cyclic-20e_nus-3d-CUSTOM')

projects/AutowareCenterPoint/datasets/tier4_dataset.py

@@ -1,14 +1,10 @@
 import os
 from os import path as osp
-from typing import Callable, List, Union
 
 import numpy as np
 
-# from .det3d_dataset import Det3DDataset
 from mmdet3d.datasets.nuscenes_dataset import NuScenesDataset
 from mmdet3d.registry import DATASETS
-from mmdet3d.structures import LiDARInstance3DBoxes
-from mmdet3d.structures.bbox_3d.cam_box3d import CameraInstance3DBoxes
 
 
 @DATASETS.register_module()

projects/AutowareCenterPoint/evaluation/__init__.py

@@ -1,3 +1,4 @@
+# flake8: noqa
 from .functional.nuscenes_utils.eval import (DetectionConfig,
                                              nuScenesDetectionEval)
 from .functional.nuscenes_utils.utils import (
@@ -6,7 +7,7 @@
 from .metrics.nuscenes_custom_metric import NuScenesCustomMetric
 
 __all__ = [
-    'NuScenesCustomMetric,'
+    'NuScenesCustomMetric'
     'DetectionConfig'
     'nuScenesDetectionEval'
     'class_mapping_kitti2nuscenes'

projects/AutowareCenterPoint/evaluation/functional/nuscenes_utils/eval.py

@@ -1,10 +1,8 @@
-import json
 import os
-from typing import Dict, List, Optional, Tuple
+from typing import Dict, List, Optional
 
 import numpy as np
 from nuscenes.eval.common.data_classes import EvalBox, EvalBoxes
-from nuscenes.eval.common.loaders import load_prediction
 from nuscenes.eval.common.utils import center_distance
 from nuscenes.eval.detection.data_classes import DetectionBox
 from nuscenes.eval.detection.evaluate import DetectionEval as _DetectionEval
@@ -57,8 +55,8 @@ def __init__(
         mean_ap_weight: int,
     ):
 
-        # assert set(class_range.keys()) == set(DETECTION_NAMES), "Class count mismatch."
-        assert dist_th_tp in dist_ths, 'dist_th_tp must be in set of dist_ths.'
+        assert dist_th_tp in dist_ths, \
+            'dist_th_tp must be in set of dist_ths.'
 
         self.class_range = class_range
         self.dist_fcn = dist_fcn
@@ -122,15 +120,19 @@ class nuScenesDetectionEval(_DetectionEval):
     written to the provided output_dir. nuScenes uses the following detection
     metrics:
 
-    - Mean Average Precision (mAP): Uses center-distance as matching criterion; averaged over distance thresholds.
-    - True Positive (TP) metrics: Average of translation, velocity, scale, orientation and attribute errors.
+    - Mean Average Precision (mAP): Uses center-distance as matching criterion;
+    averaged over distance thresholds.
+    - True Positive (TP) metrics: Average of translation, velocity, scale,
+     orientation and attribute errors.
     - nuScenes Detection Score (NDS): The weighted sum of the above.
     Here is an overview of the functions in this method:
-    - init: Loads GT annotations and predictions stored in JSON format and filters the boxes.
+    - init: Loads GT annotations and predictions stored in
+    JSON format and filters the boxes.
     - run: Performs evaluation and dumps the metric data to disk.
     - render: Renders various plots and dumps to disk.
     We assume that:
-    - Every sample_token is given in the results, although there may be not predictions for that sample.
+    - Every sample_token is given in the results, although there may be
+     not predictions for that sample.
     Please see https://www.nuscenes.org/object-detection for more details.
     """
 
@@ -149,7 +151,8 @@ def __init__(
         :param config: A DetectionConfig object.
         :param result_boxes: result bounding boxes.
         :param gt_boxes: ground-truth bounding boxes.
-        :param eval_set: The dataset split to evaluate on, e.g. train, val or test.
+        :param eval_set: The dataset split to evaluate on,
+        e.g. train, val or test.
         :param output_dir: Folder to save plots and results to.
         :param verbose: Whether to print to stdout.
         """

projects/AutowareCenterPoint/evaluation/functional/nuscenes_utils/utils.py

@@ -234,7 +234,7 @@ def format_nuscenes_metrics(metrics: Dict,
     result_dict: Dict[str, Dict[str, float]] = defaultdict(dict)
     for name in class_names:
         result_dict[name].update({
-            f'mAP':
+            'mAP':
             sum([v for v in metrics['label_aps'][name].values()]) * 100 /
             len(metrics['label_aps'][name])
         })

projects/AutowareCenterPoint/evaluation/metrics/nuscenes_custom_metric.py

@@ -3,7 +3,6 @@
 from os import path as osp
 from typing import Dict, List, Optional, Sequence, Tuple, Union
 
-import mmcv
 import mmengine
 import numpy as np
 import pyquaternion
@@ -235,8 +234,6 @@ def _evaluate_single(
         Returns:
             Dict[str, float]: Dictionary of evaluation details.
         """
-        from nuscenes import NuScenes
-        from nuscenes.eval.detection.evaluate import NuScenesEval
 
         output_dir = osp.join(*osp.split(result_path)[:-1])
 

tools/dataset_converters/tier4dataset_converter.py

@@ -216,7 +216,7 @@ def _fill_trainval_infos(nusc,
 
         for cam in camera_types:
 
-            if not cam in sample['data']:
+            if cam not in sample['data']:
                 continue
 
             cam_token = sample['data'][cam]
@@ -440,9 +440,10 @@ def get_2d_boxes(nusc,
     # Get the sample data and the sample corresponding to that sample data.
     sd_rec = nusc.get('sample_data', sample_data_token)
 
-    assert sd_rec[
-               'sensor_modality'] == 'camera', 'Error: get_2d_boxes only works' \
-                                               ' for camera sample_data!'
+    assert sd_rec['sensor_modality'] == 'camera', ('Error: get_2d_boxes only '
+                                                   'worksfor camera '
+                                                   'sample_data!')
+
     if not sd_rec['is_key_frame']:
         raise ValueError(
             'The 2D re-projections are available only for keyframes.')