[Fix] replace mmcv's function and modules imported with mmengine's and sync...

[Fix] replace mmcv's function and modules imported with mmengine's and sync the latest mmengine (#1765)

* replace mmcv to mmengine

* fix

* fix comments

.dev_scripts/gather_models.py

@@ -22,7 +22,6 @@ import shutil
 import subprocess
 from os import path as osp
 
-import mmcv
 import mmengine
 import torch
 
@@ -143,10 +142,10 @@ def main():
     args = parse_args()
     models_root = args.root
     models_out = args.out
-    mmcv.mkdir_or_exist(models_out)
+    mmengine.mkdir_or_exist(models_out)
 
     # find all models in the root directory to be gathered
-    raw_configs = list(mmcv.scandir('./configs', '.py', recursive=True))
+    raw_configs = list(mmengine.scandir('./configs', '.py', recursive=True))
 
     # filter configs that is not trained in the experiments dir
     used_configs = []
@@ -189,7 +188,7 @@ def main():
     publish_model_infos = []
     for model in model_infos:
         model_publish_dir = osp.join(models_out, model['config'].rstrip('.py'))
-        mmcv.mkdir_or_exist(model_publish_dir)
+        mmengine.mkdir_or_exist(model_publish_dir)
 
         model_name = model['config'].split('/')[-1].rstrip(
             '.py') + '_' + model['model_time']

data/s3dis/collect_indoor3d_data.py

 import argparse
 from os import path as osp
 
-import mmcv
+import mmengine
 from indoor3d_util import export
 
 parser = argparse.ArgumentParser()
@@ -23,7 +23,7 @@ anno_paths = [line.rstrip() for line in open(args.ann_file)]
 anno_paths = [osp.join(args.data_dir, p) for p in anno_paths]
 
 output_folder = args.output_folder
-mmcv.mkdir_or_exist(output_folder)
+mmengine.mkdir_or_exist(output_folder)
 
 # Note: there is an extra character in the v1.2 data in Area_5/hallway_6.
 # It's fixed manually here.

data/scannet/extract_posed_images.py

@@ -6,7 +6,7 @@ from argparse import ArgumentParser
 from functools import partial
 
 import imageio
-import mmcv
+import mmengine
 import numpy as np
 
 COMPRESSION_TYPE_COLOR = {-1: 'unknown', 0: 'raw', 1: 'png', 2: 'jpeg'}
@@ -160,7 +160,7 @@ def process_scene(path, limit, idx):
 
 def process_directory(path, limit, nproc):
     print(f'processing {path}')
-    mmcv.track_parallel_progress(
+    mmengine.track_parallel_progress(
         func=partial(process_scene, path, limit),
         tasks=os.listdir(path),
         nproc=nproc)

docs/en/advanced_guides/datasets/sunrgbd_det.md

@@ -150,7 +150,7 @@ def process_single_scene(sample_idx):
     info['point_cloud'] = pc_info
 
     # save point cloud data in `.bin` format
-    mmcv.mkdir_or_exist(osp.join(self.root_dir, 'points'))
+    mmengine.mkdir_or_exist(osp.join(self.root_dir, 'points'))
     pc_upright_depth_subsampled.tofile(
         osp.join(self.root_dir, 'points', f'{sample_idx:06d}.bin'))
 

docs/zh_cn/advanced_guides/datasets/sunrgbd_det.md

@@ -148,7 +148,7 @@ def process_single_scene(sample_idx):
     info['point_cloud'] = pc_info
 
     # 将点云保存为 `.bin` 格式
-    mmcv.mkdir_or_exist(osp.join(self.root_dir, 'points'))
+    mmengine.mkdir_or_exist(osp.join(self.root_dir, 'points'))
     pc_upright_depth_subsampled.tofile(
         osp.join(self.root_dir, 'points', f'{sample_idx:06d}.bin'))
 

mmdet3d/apis/inference.py

@@ -22,7 +22,7 @@ def convert_SyncBN(config):
     """Convert config's naiveSyncBN to BN.
 
     Args:
-         config (str or :obj:`mmcv.Config`): Config file path or the config
+         config (str or :obj:`mmengine.Config`): Config file path or the config
             object.
     """
     if isinstance(config, dict):

mmdet3d/datasets/transforms/formating.py

 # Copyright (c) OpenMMLab. All rights reserved.
 from typing import List, Sequence, Union
 
-import mmcv
+import mmengine
 import numpy as np
 import torch
 from mmcv import BaseTransform
-from mmengine import InstanceData
+from mmengine.structures import InstanceData
 from numpy import dtype
 
 from mmdet3d.registry import TRANSFORMS
@@ -35,7 +35,7 @@ def to_tensor(
         if data.dtype is dtype('float64'):
             data = data.astype(np.float32)
         return torch.from_numpy(data)
-    elif isinstance(data, Sequence) and not mmcv.is_str(data):
+    elif isinstance(data, Sequence) and not mmengine.is_str(data):
         return torch.tensor(data)
     elif isinstance(data, int):
         return torch.LongTensor([data])

mmdet3d/datasets/transforms/loading.py

@@ -182,7 +182,7 @@ class LoadPointsFromMultiSweeps(BaseTransform):
             pts_bytes = self.file_client.get(pts_filename)
             points = np.frombuffer(pts_bytes, dtype=np.float32)
         except ConnectionError:
-            mmcv.check_file_exist(pts_filename)
+            mmengine.check_file_exist(pts_filename)
             if pts_filename.endswith('.npy'):
                 points = np.load(pts_filename)
             else:
@@ -440,7 +440,7 @@ class LoadPointsFromFile(BaseTransform):
             pts_bytes = self.file_client.get(pts_filename)
             points = np.frombuffer(pts_bytes, dtype=np.float32)
         except ConnectionError:
-            mmcv.check_file_exist(pts_filename)
+            mmengine.check_file_exist(pts_filename)
             if pts_filename.endswith('.npy'):
                 points = np.load(pts_filename)
             else:
@@ -696,7 +696,7 @@ class LoadAnnotations3D(LoadAnnotations):
             mask_bytes = self.file_client.get(pts_instance_mask_path)
             pts_instance_mask = np.frombuffer(mask_bytes, dtype=np.int64)
         except ConnectionError:
-            mmcv.check_file_exist(pts_instance_mask_path)
+            mmengine.check_file_exist(pts_instance_mask_path)
             pts_instance_mask = np.fromfile(
                 pts_instance_mask_path, dtype=np.int64)
 
@@ -725,7 +725,7 @@ class LoadAnnotations3D(LoadAnnotations):
             pts_semantic_mask = np.frombuffer(
                 mask_bytes, dtype=self.seg_3d_dtype).copy()
         except ConnectionError:
-            mmcv.check_file_exist(pts_semantic_mask_path)
+            mmengine.check_file_exist(pts_semantic_mask_path)
             pts_semantic_mask = np.fromfile(
                 pts_semantic_mask_path, dtype=np.int64)
 

mmdet3d/datasets/transforms/test_time_aug.py

@@ -3,7 +3,7 @@ import warnings
 from copy import deepcopy
 from typing import Dict, List, Optional, Tuple, Union
 
-import mmcv
+import mmengine
 from mmcv import BaseTransform
 from mmengine.dataset import Compose
 
@@ -48,8 +48,8 @@ class MultiScaleFlipAug3D(BaseTransform):
         self.pts_scale_ratio = pts_scale_ratio \
             if isinstance(pts_scale_ratio, list) else[float(pts_scale_ratio)]
 
-        assert mmcv.is_list_of(self.img_scale, tuple)
-        assert mmcv.is_list_of(self.pts_scale_ratio, float)
+        assert mmengine.is_list_of(self.img_scale, tuple)
+        assert mmengine.is_list_of(self.pts_scale_ratio, float)
 
         self.flip = flip
         self.pcd_horizontal_flip = pcd_horizontal_flip
@@ -57,7 +57,7 @@ class MultiScaleFlipAug3D(BaseTransform):
 
         self.flip_direction = flip_direction if isinstance(
             flip_direction, list) else [flip_direction]
-        assert mmcv.is_list_of(self.flip_direction, str)
+        assert mmengine.is_list_of(self.flip_direction, str)
         if not self.flip and self.flip_direction != ['horizontal']:
             warnings.warn(
                 'flip_direction has no effect when flip is set to False')

mmdet3d/datasets/transforms/transforms_3d.py

@@ -5,8 +5,8 @@ from typing import Dict, List, Optional, Tuple, Union
 
 import cv2
 import numpy as np
-from mmcv import is_tuple_of
 from mmcv.transforms import BaseTransform
+from mmengine import is_tuple_of
 
 from mmdet3d.models.task_modules import VoxelGenerator
 from mmdet3d.registry import TRANSFORMS

mmdet3d/datasets/utils.py

 # Copyright (c) OpenMMLab. All rights reserved.
-import mmcv
 import numpy as np
 from mmcv.transforms import LoadImageFromFile
 from pyquaternion import Quaternion
@@ -113,33 +112,6 @@ def get_loading_pipeline(pipeline):
     return loading_pipeline
 
 
-def extract_result_dict(results, key):
-    """Extract and return the data corresponding to key in result dict.
-
-    ``results`` is a dict output from `pipeline(input_dict)`, which is the
-        loaded data from ``Dataset`` class.
-    The data terms inside may be wrapped in list, tuple and DataContainer, so
-        this function essentially extracts data from these wrappers.
-
-    Args:
-        results (dict): Data loaded using pipeline.
-        key (str): Key of the desired data.
-
-    Returns:
-        np.ndarray | torch.Tensor: Data term.
-    """
-    if key not in results.keys():
-        return None
-    # results[key] may be data or list[data] or tuple[data]
-    # data may be wrapped inside DataContainer
-    data = results[key]
-    if isinstance(data, (list, tuple)):
-        data = data[0]
-    if isinstance(data, mmcv.parallel.DataContainer):
-        data = data._data
-    return data
-
-
 def convert_quaternion_to_matrix(quaternion: list,
                                  translation: list = None) -> list:
     """Compute a transform matrix by given quaternion and translation

mmdet3d/evaluation/functional/indoor_eval.py

 # Copyright (c) OpenMMLab. All rights reserved.
 import numpy as np
 import torch
-from mmcv.utils import print_log
+from mmengine.logging import print_log
 from terminaltables import AsciiTable
 
 

mmdet3d/evaluation/functional/instance_seg_eval.py

 # Copyright (c) OpenMMLab. All rights reserved.
 import numpy as np
-from mmcv.utils import print_log
+from mmengine.logging import print_log
 from terminaltables import AsciiTable
 
 from .scannet_utils.evaluate_semantic_instance import scannet_eval

mmdet3d/evaluation/functional/lyft_eval.py

 # Copyright (c) OpenMMLab. All rights reserved.
 from os import path as osp
 
-import mmcv
 import mmengine
 import numpy as np
 from lyft_dataset_sdk.eval.detection.mAP_evaluation import (Box3D, get_ap,
@@ -9,7 +8,7 @@ from lyft_dataset_sdk.eval.detection.mAP_evaluation import (Box3D, get_ap,
                                                             get_ious,
                                                             group_by_key,
                                                             wrap_in_box)
-from mmcv.utils import print_log
+from mmengine.logging import print_log
 from terminaltables import AsciiTable
 
 
@@ -26,7 +25,7 @@ def load_lyft_gts(lyft, data_root, eval_split, logger=None):
     Returns:
         list[dict]: List of annotation dictionaries.
     """
-    split_scenes = mmcv.list_from_file(
+    split_scenes = mmengine.list_from_file(
         osp.join(data_root, f'{eval_split}.txt'))
 
     # Read out all sample_tokens in DB.
@@ -50,7 +49,7 @@ def load_lyft_gts(lyft, data_root, eval_split, logger=None):
 
     print_log('Loading ground truth annotations...', logger=logger)
     # Load annotations and filter predictions and annotations.
-    for sample_token in mmcv.track_iter_progress(sample_tokens):
+    for sample_token in mmengine.track_iter_progress(sample_tokens):
         sample = lyft.get('sample', sample_token)
         sample_annotation_tokens = sample['anns']
         for sample_annotation_token in sample_annotation_tokens:

mmdet3d/evaluation/functional/seg_eval.py

 # Copyright (c) OpenMMLab. All rights reserved.
 import numpy as np
-from mmcv.utils import print_log
+from mmengine.logging import print_log
 from terminaltables import AsciiTable
 
 

mmdet3d/evaluation/functional/waymo_utils/prediction_kitti_to_waymo.py

@@ -13,7 +13,7 @@ except ImportError:
 from glob import glob
 from os.path import join
 
-import mmcv
+import mmengine
 import numpy as np
 import tensorflow as tf
 from waymo_open_dataset import label_pb2
@@ -101,7 +101,7 @@ class KITTI2Waymo(object):
 
     def create_folder(self):
         """Create folder for data conversion."""
-        mmcv.mkdir_or_exist(self.waymo_results_save_dir)
+        mmengine.mkdir_or_exist(self.waymo_results_save_dir)
 
     def parse_objects(self, kitti_result, T_k2w, context_name,
                       frame_timestamp_micros):
@@ -228,8 +228,8 @@ class KITTI2Waymo(object):
     def convert(self):
         """Convert action."""
         print('Start converting ...')
-        mmcv.track_parallel_progress(self.convert_one, range(len(self)),
-                                     self.workers)
+        mmengine.track_parallel_progress(self.convert_one, range(len(self)),
+                                         self.workers)
         print('\nFinished ...')
 
         # combine all files into one .bin

mmdet3d/evaluation/metrics/kitti_metric.py

@@ -3,14 +3,12 @@ import tempfile
 from os import path as osp
 from typing import Dict, List, Optional, Sequence, Union
 
-import mmcv
 import mmengine
 import numpy as np
 import torch
-from mmcv.utils import print_log
 from mmengine import load
 from mmengine.evaluator import BaseMetric
-from mmengine.logging import MMLogger
+from mmengine.logging import MMLogger, print_log
 
 from mmdet3d.evaluation import kitti_eval
 from mmdet3d.registry import METRICS
@@ -328,12 +326,12 @@ class KittiMetric(BaseMetric):
         assert len(net_outputs) == len(self.data_infos), \
             'invalid list length of network outputs'
         if submission_prefix is not None:
-            mmcv.mkdir_or_exist(submission_prefix)
+            mmengine.mkdir_or_exist(submission_prefix)
 
         det_annos = []
         print('\nConverting prediction to KITTI format')
         for idx, pred_dicts in enumerate(
-                mmcv.track_iter_progress(net_outputs)):
+                mmengine.track_iter_progress(net_outputs)):
             annos = []
             sample_idx = sample_id_list[idx]
             info = self.data_infos[sample_idx]
@@ -459,7 +457,7 @@ class KittiMetric(BaseMetric):
         det_annos = []
         print('\nConverting prediction to KITTI format')
         for i, bboxes_per_sample in enumerate(
-                mmcv.track_iter_progress(net_outputs)):
+                mmengine.track_iter_progress(net_outputs)):
             annos = []
             anno = dict(
                 name=[],
@@ -523,7 +521,7 @@ class KittiMetric(BaseMetric):
 
         if submission_prefix is not None:
             # save file in submission format
-            mmcv.mkdir_or_exist(submission_prefix)
+            mmengine.mkdir_or_exist(submission_prefix)
             print(f'Saving KITTI submission to {submission_prefix}')
             for i, anno in enumerate(det_annos):
                 sample_idx = self.data_infos[i]['image']['image_idx']

mmdet3d/evaluation/metrics/lyft_metric.py

@@ -5,7 +5,6 @@ import tempfile
 from os import path as osp
 from typing import Dict, List, Optional, Sequence, Tuple, Union
 
-import mmcv
 import mmengine
 import numpy as np
 import pandas as pd
@@ -220,7 +219,7 @@ class LyftMetric(BaseMetric):
             idx = Id_list.index(token)
             pred_list[idx] = prediction_str
         df = pd.DataFrame({'Id': Id_list, 'PredictionString': pred_list})
-        mmcv.mkdir_or_exist(os.path.dirname(csv_savepath))
+        mmengine.mkdir_or_exist(os.path.dirname(csv_savepath))
         df.to_csv(csv_savepath, index=False)
 
     def _format_bbox(self,
@@ -245,7 +244,7 @@ class LyftMetric(BaseMetric):
         lyft_annos = {}
 
         print('Start to convert detection format...')
-        for i, det in enumerate(mmcv.track_iter_progress(results)):
+        for i, det in enumerate(mmengine.track_iter_progress(results)):
             annos = []
             boxes = output_to_lyft_box(det)
             sample_id = sample_id_list[i]
@@ -267,7 +266,7 @@ class LyftMetric(BaseMetric):
             'results': lyft_annos,
         }
 
-        mmcv.mkdir_or_exist(jsonfile_prefix)
+        mmengine.mkdir_or_exist(jsonfile_prefix)
         res_path = osp.join(jsonfile_prefix, 'results_lyft.json')
         print('Results writes to', res_path)
         mmengine.dump(lyft_submissions, res_path)

mmdet3d/evaluation/metrics/nuscenes_metric.py

@@ -4,12 +4,11 @@ import tempfile
 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
 import torch
-from mmengine import load
+from mmengine import Config, load
 from mmengine.evaluator import BaseMetric
 from mmengine.logging import MMLogger
 from nuscenes.eval.detection.config import config_factory
@@ -379,7 +378,7 @@ class NuScenesMetric(BaseMetric):
 
         CAM_NUM = 6
 
-        for i, det in enumerate(mmcv.track_iter_progress(results)):
+        for i, det in enumerate(mmengine.track_iter_progress(results)):
 
             sample_id = sample_id_list[i]
 
@@ -417,7 +416,6 @@ class NuScenesMetric(BaseMetric):
                 score_thr=0.01,
                 min_bbox_size=0,
                 max_per_frame=500)
-            from mmcv import Config
             nms_cfg = Config(nms_cfg)
             cam_boxes3d_for_nms = xywhr2xyxyr(cam_boxes3d.bev)
             boxes3d = cam_boxes3d.tensor
@@ -462,7 +460,7 @@ class NuScenesMetric(BaseMetric):
             'results': nusc_annos,
         }
 
-        mmcv.mkdir_or_exist(jsonfile_prefix)
+        mmengine.mkdir_or_exist(jsonfile_prefix)
         res_path = osp.join(jsonfile_prefix, 'results_nusc.json')
         print('Results writes to', res_path)
         mmengine.dump(nusc_submissions, res_path)
@@ -490,7 +488,7 @@ class NuScenesMetric(BaseMetric):
         nusc_annos = {}
 
         print('Start to convert detection format...')
-        for i, det in enumerate(mmcv.track_iter_progress(results)):
+        for i, det in enumerate(mmengine.track_iter_progress(results)):
             annos = []
             boxes = output_to_nusc_box(det)
             sample_id = sample_id_list[i]
@@ -536,7 +534,7 @@ class NuScenesMetric(BaseMetric):
             'meta': self.modality,
             'results': nusc_annos,
         }
-        mmcv.mkdir_or_exist(jsonfile_prefix)
+        mmengine.mkdir_or_exist(jsonfile_prefix)
         res_path = osp.join(jsonfile_prefix, 'results_nusc.json')
         print('Results writes to', res_path)
         mmengine.dump(nusc_submissions, res_path)

mmdet3d/evaluation/metrics/waymo_metric.py

@@ -3,13 +3,11 @@ import tempfile
 from os import path as osp
 from typing import Dict, List, Optional, Union
 
-import mmcv
 import mmengine
 import numpy as np
 import torch
-from mmcv.utils import print_log
-from mmengine import load
-from mmengine.logging import MMLogger
+from mmengine import Config, load
+from mmengine.logging import MMLogger, print_log
 
 from mmdet3d.models.layers import box3d_multiclass_nms
 from mmdet3d.registry import METRICS
@@ -271,7 +269,6 @@ class WaymoMetric(KittiMetric):
             score_thr=0.001,
             min_bbox_size=0,
             max_per_frame=100)
-        from mmcv import Config
         nms_cfg = Config(nms_cfg)
         lidar_boxes3d = LiDARInstance3DBoxes(
             torch.from_numpy(box_dict['box3d_lidar']).cuda())
@@ -332,12 +329,12 @@ class WaymoMetric(KittiMetric):
         assert len(net_outputs) == len(self.data_infos), \
             'invalid list length of network outputs'
         if submission_prefix is not None:
-            mmcv.mkdir_or_exist(submission_prefix)
+            mmengine.mkdir_or_exist(submission_prefix)
 
         det_annos = []
         print('\nConverting prediction to KITTI format')
         for idx, pred_dicts in enumerate(
-                mmcv.track_iter_progress(net_outputs)):
+                mmengine.track_iter_progress(net_outputs)):
             annos = []
             sample_idx = sample_id_list[idx]
             info = self.data_infos[sample_idx]