Merge 3b0e12d into af88d62

mmpose/apis/inferencers/pose2d_inferencer.py

@@ -4,7 +4,6 @@
 
 import mmcv
 import numpy as np
-from mmdet.apis.det_inferencer import DetInferencer
 from mmengine.config import Config, ConfigDict
 from mmengine.infer.infer import ModelType
 from mmengine.registry import init_default_scope
@@ -17,6 +16,12 @@
 from .base_mmpose_inferencer import BaseMMPoseInferencer
 from .utils import default_det_models
 
+try:
+    from mmdet.apis.det_inferencer import DetInferencer
+    mmdet_available = True
+except (ImportError, ModuleNotFoundError):
+    mmdet_available = False
+
 InstanceList = List[InstanceData]
 InputType = Union[str, np.ndarray]
 InputsType = Union[InputType, Sequence[InputType]]
@@ -94,8 +99,14 @@ def __init__(self,
                 det_model, det_weights, det_cat_ids = det_info[
                     'model'], det_info['weights'], det_info['cat_ids']
 
-            self.detector = DetInferencer(
-                det_model, det_weights, device=device)
+            if mmdet_available:
+                self.detector = DetInferencer(
+                    det_model, det_weights, device=device)
+            else:
+                raise RuntimeError(
+                    'MMDetection is required to build inferencers '
+                    'for top-down pose estimation models.')
+
             if isinstance(det_cat_ids, (tuple, list)):
                 self.det_cat_ids = det_cat_ids
             else:

mmpose/models/pose_estimators/base.py

@@ -90,7 +90,7 @@ def _load_metainfo(metainfo: dict = None) -> dict:
 
     def forward(self,
                 inputs: torch.Tensor,
-                data_samples: OptSampleList,
+                data_samples: OptSampleList = None,
                 mode: str = 'tensor') -> ForwardResults:
         """The unified entry for a forward process in both training and test.
 

tools/analysis_tools/get_flops.py

@@ -1,111 +1,81 @@
 # Copyright (c) OpenMMLab. All rights reserved.
 import argparse
-from functools import partial
 
-import torch
-from mmengine.config import DictAction
+from mmengine.config import Config, DictAction
 
-from mmpose.apis.inference import init_model
+from mmpose.registry import MODELS
+from mmpose.utils import register_all_modules
 
 try:
-    from mmcv.cnn import get_model_complexity_info
+    from mmengine.analysis import get_model_complexity_info
 except ImportError:
     raise ImportError('Please upgrade mmcv to >0.6.2')
 
 
 def parse_args():
-    parser = argparse.ArgumentParser(description='Train a recognizer')
+    parser = argparse.ArgumentParser(description='Train a detector')
     parser.add_argument('config', help='train config file path')
-    parser.add_argument(
-        '--device',
-        default='cuda:0',
-        help='Device used for model initialization')
-    parser.add_argument(
-        '--cfg-options',
-        nargs='+',
-        action=DictAction,
-        default={},
-        help='override some settings in the used config, the key-value pair '
-        'in xxx=yyy format will be merged into config file. For example, '
-        "'--cfg-options model.backbone.depth=18 model.backbone.with_cp=True'")
     parser.add_argument(
         '--shape',
         type=int,
         nargs='+',
-        default=[256, 192],
+        default=[1280, 800],
         help='input image size')
     parser.add_argument(
-        '--input-constructor',
-        '-c',
-        type=str,
-        choices=['none', 'batch'],
-        default='none',
-        help='If specified, it takes a callable method that generates '
-        'input. Otherwise, it will generate a random tensor with '
-        'input shape to calculate FLOPs.')
-    parser.add_argument(
-        '--batch-size', '-b', type=int, default=1, help='input batch size')
-    parser.add_argument(
-        '--not-print-per-layer-stat',
-        '-n',
-        action='store_true',
-        help='Whether to print complexity information'
-        'for each layer in a model')
+        '--cfg-options',
+        nargs='+',
+        action=DictAction,
+        help='override some settings in the used config, the key-value pair '
+        'in xxx=yyy format will be merged into config file. If the value to '
+        'be overwritten is a list, it should be like key="[a,b]" or key=a,b '
+        'It also allows nested list/tuple values, e.g. key="[(a,b),(c,d)]" '
+        'Note that the quotation marks are necessary and that no white space '
+        'is allowed.')
     args = parser.parse_args()
     return args
 
 
-def batch_constructor(flops_model, batch_size, input_shape):
-    """Generate a batch of tensors to the model."""
-    batch = {}
-
-    inputs = torch.ones(()).new_empty(
-        (batch_size, *input_shape),
-        dtype=next(flops_model.parameters()).dtype,
-        device=next(flops_model.parameters()).device)
-
-    batch['inputs'] = inputs
-    return batch
-
-
 def main():
-
+    register_all_modules()
     args = parse_args()
 
     if len(args.shape) == 1:
-        input_shape = (3, args.shape[0], args.shape[0])
+        h = w = args.shape[0]
     elif len(args.shape) == 2:
-        input_shape = (3, ) + tuple(args.shape)
+        h, w = args.shape
     else:
         raise ValueError('invalid input shape')
+    input_shape = (3, h, w)
 
-    model = init_model(
-        args.config,
-        checkpoint=None,
-        device=args.device,
-        cfg_options=args.cfg_options)
+    cfg = Config.fromfile(args.config)
+    if args.cfg_options is not None:
+        cfg.merge_from_dict(args.cfg_options)
 
-    if args.input_constructor == 'batch':
-        input_constructor = partial(batch_constructor, model, args.batch_size)
-    else:
-        input_constructor = None
+    model = MODELS.build(cfg.model)
+    model.eval()
+
+    analysis_results = get_model_complexity_info(
+        model, input_shape, show_table=True, show_arch=False)
+
+    # ayalysis_results = {
+    #     'flops': flops,
+    #     'flops_str': flops_str,
+    #     'activations': activations,
+    #     'activations_str': activations_str,
+    #     'params': params,
+    #     'params_str': params_str,
+    #     'out_table': complexity_table,
+    #     'out_arch': complexity_arch
+    # }
 
-    if hasattr(model, '_forward'):
-        model.forward = model._forward
-    else:
-        raise NotImplementedError(
-            'FLOPs counter is currently not currently supported with {}'.
-            format(model.__class__.__name__))
-
-    flops, params = get_model_complexity_info(
-        model,
-        input_shape,
-        input_constructor=input_constructor,
-        print_per_layer_stat=(not args.not_print_per_layer_stat))
     split_line = '=' * 30
-    input_shape = (args.batch_size, ) + input_shape
     print(f'{split_line}\nInput shape: {input_shape}\n'
-          f'Flops: {flops}\nParams: {params}\n{split_line}')
+          f'Flops: {analysis_results["flops"]}\n'
+          f'Params: {analysis_results["params"]}\n{split_line}')
+
+    print(analysis_results['activations'])
+    # print(analysis_results['complexity_table'])
+    # print(complexity_str)
     print('!!!Please be cautious if you use the results in papers. '
           'You may need to check if all ops are supported and verify that the '
           'flops computation is correct.')

tools/analysis_tools/get_flops1.py

@@ -0,0 +1,106 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+import argparse
+
+import torch
+from mmengine.config import DictAction
+
+from mmpose.apis.inference import init_model
+
+try:
+    # from mmcv.cnn import get_model_complexity_info
+    from mmengine.analysis import get_model_complexity_info
+except ImportError:
+    raise ImportError('Please upgrade mmcv to >0.6.2')
+
+
+def parse_args():
+    parser = argparse.ArgumentParser(description='Train a recognizer')
+    parser.add_argument('config', help='train config file path')
+    parser.add_argument(
+        '--device', default='cpu', help='Device used for model initialization')
+    parser.add_argument(
+        '--cfg-options',
+        nargs='+',
+        action=DictAction,
+        default={},
+        help='override some settings in the used config, the key-value pair '
+        'in xxx=yyy format will be merged into config file. For example, '
+        "'--cfg-options model.backbone.depth=18 model.backbone.with_cp=True'")
+    parser.add_argument(
+        '--shape',
+        type=int,
+        nargs='+',
+        default=[256, 192],
+        help='input image size')
+    parser.add_argument(
+        '--input-constructor',
+        '-c',
+        type=str,
+        choices=['none', 'batch'],
+        default='none',
+        help='If specified, it takes a callable method that generates '
+        'input. Otherwise, it will generate a random tensor with '
+        'input shape to calculate FLOPs.')
+    parser.add_argument(
+        '--batch-size', '-b', type=int, default=1, help='input batch size')
+    parser.add_argument(
+        '--not-print-per-layer-stat',
+        '-n',
+        action='store_true',
+        help='Whether to print complexity information'
+        'for each layer in a model')
+    args = parser.parse_args()
+    return args
+
+
+def batch_constructor(flops_model, batch_size, input_shape):
+    """Generate a batch of tensors to the model."""
+    batch = {}
+
+    inputs = torch.ones(()).new_empty(
+        (batch_size, *input_shape),
+        dtype=next(flops_model.parameters()).dtype,
+        device=next(flops_model.parameters()).device)
+
+    batch['inputs'] = inputs
+    return batch
+
+
+def main():
+
+    args = parse_args()
+
+    if len(args.shape) == 1:
+        input_shape = (3, args.shape[0], args.shape[0])
+    elif len(args.shape) == 2:
+        input_shape = (3, ) + tuple(args.shape)
+    else:
+        raise ValueError('invalid input shape')
+
+    model = init_model(
+        args.config,
+        checkpoint=None,
+        device=args.device,
+        cfg_options=args.cfg_options)
+
+    if hasattr(model, '_forward'):
+        model.forward = model._forward
+    else:
+        raise NotImplementedError(
+            'FLOPs counter is currently not currently supported with {}'.
+            format(model.__class__.__name__))
+
+    analysis_results = get_model_complexity_info(model, input_shape)
+    flops = analysis_results['flops_str']
+    params = analysis_results['params_str']
+    split_line = '=' * 30
+    input_shape = (args.batch_size, ) + input_shape
+    print(f'{split_line}\nInput shape: {input_shape}\n'
+          f'Flops: {flops}\nParams: {params}\n{split_line}')
+    print('!!!Please be cautious if you use the results in papers. '
+          'You may need to check if all ops are supported and verify that the '
+          'flops computation is correct.')
+
+
+if __name__ == '__main__':
+    main()