def eval_results(results,
feed,
metric,
num_classes,
resolution=None,
is_bbox_normalized=False,
output_file=None):
"""Evaluation for evaluation program results"""
if metric == 'COCO':
from ppdet.utils.coco_eval import proposal_eval, bbox_eval, mask_eval
anno_file = getattr(feed.dataset, 'annotation', None)
with_background = getattr(feed, 'with_background', True)
if 'proposal' in results[0]:
output = 'proposal.json'
if output_file:
output = '{}_proposal.json'.format(output_file)
proposal_eval(results, anno_file, output)
if 'bbox' in results[0]:
output = 'bbox.json'
if output_file:
output = '{}_bbox.json'.format(output_file)
bbox_eval(results, anno_file, output, with_background)
if 'mask' in results[0]:
output = 'mask.json'
if output_file:
output = '{}_mask.json'.format(output_file)
mask_eval(results, anno_file, output, resolution)
else:
if 'accum_map' in results[-1]:
res = np.mean(results[-1]['accum_map'][0])
logger.info('mAP: {:.2f}'.format(res * 100.))
elif 'bbox' in results[0]:
voc_bbox_eval(results,
num_classes,
is_bbox_normalized=is_bbox_normalized)
def eval_results(results, feed, metric, resolution=None, output_file=None):
"""Evaluation for evaluation program results"""
if metric == 'COCO':
from ppdet.utils.coco_eval import bbox_eval, mask_eval
anno_file = getattr(feed.dataset, 'annotation', None)
with_background = getattr(feed, 'with_background', True)
output = 'bbox.json'
if output_file:
output = '{}_bbox.json'.format(output_file)
bbox_eval(results, anno_file, output, with_background)
if 'mask' in results[0]:
output = 'mask.json'
if output_file:
output = '{}_mask.json'.format(output_file)
mask_eval(results, anno_file, output, resolution)
else:
res = np.mean(results[-1]['accum_map'][0])
logger.info('Test mAP: {}'.format(res))
def eval_results(results,
metric,
num_classes,
resolution=None,
is_bbox_normalized=False,
output_directory=None,
map_type='11point',
dataset=None,
save_only=False,
test_scales=None):
"""Evaluation for evaluation program results"""
box_ap_stats = []
if metric == 'COCO':
from ppdet.utils.coco_eval import proposal_eval, bbox_eval, mask_eval
if output_directory:
mmcv.mkdir_or_exist(output_directory)
anno_file = dataset.get_anno()
with_background = dataset.with_background
if 'proposal' in results[0]:
output = 'proposal.json'
if output_directory:
output = os.path.join(output_directory, 'proposal.json')
proposal_eval(results, anno_file, output)
if 'bbox' in results[0]:
output = 'bbox_scale{}.json'.format(test_scales) \
if test_scales else 'bbox.json'
if output_directory:
output = os.path.join(output_directory, output)
box_ap_stats = bbox_eval(results,
anno_file,
output,
with_background,
is_bbox_normalized=is_bbox_normalized,
save_only=save_only)
if 'mask' in results[0]:
output = 'mask.json'
if output_directory:
output = os.path.join(output_directory, 'mask.json')
mask_eval(results,
anno_file,
output,
resolution,
save_only=save_only)
else:
if 'accum_map' in results[-1]:
res = np.mean(results[-1]['accum_map'][0])
logger.info('mAP: {:.2f}'.format(res * 100.))
box_ap_stats.append(res * 100.)
elif 'bbox' in results[0]:
box_ap = voc_bbox_eval(results,
num_classes,
is_bbox_normalized=is_bbox_normalized,
map_type=map_type)
box_ap_stats.append(box_ap)
return box_ap_stats
def eval_results(metric,
dataset=None):
"""Evaluation for evaluation program results"""
box_ap_stats = []
if metric == 'COCO':
from ppdet.utils.coco_eval import proposal_eval, bbox_eval, mask_eval
anno_file = dataset.get_anno()
box_ap_stats = bbox_eval(anno_file)
mask_ap_stats = mask_eval(anno_file)
else:
pass
return box_ap_stats, mask_ap_stats
def eval_results(results,
feed,
metric,
num_classes,
resolution=None,
is_bbox_normalized=False,
output_directory=None,
map_type='11point'):
"""Evaluation for evaluation program results"""
box_ap_stats = []
if metric == 'COCO':
from ppdet.utils.coco_eval import proposal_eval, bbox_eval, mask_eval
anno_file = getattr(feed.dataset, 'annotation', None)
with_background = getattr(feed, 'with_background', True)
if 'proposal' in results[0]:
output = 'proposal.json'
if output_directory:
output = os.path.join(output_directory, 'proposal.json')
proposal_eval(results, anno_file, output)
if 'bbox' in results[0]:
output = 'bbox.json'
if output_directory:
output = os.path.join(output_directory, 'bbox.json')
box_ap_stats = bbox_eval(results,
anno_file,
output,
with_background,
is_bbox_normalized=is_bbox_normalized)
if 'mask' in results[0]:
output = 'mask.json'
if output_directory:
output = os.path.join(output_directory, 'mask.json')
mask_eval(results, anno_file, output, resolution)
else:
if 'accum_map' in results[-1]:
res = np.mean(results[-1]['accum_map'][0])
logger.info('mAP: {:.2f}'.format(res * 100.))
box_ap_stats.append(res * 100.)
elif 'bbox' in results[0]:
box_ap = voc_bbox_eval(results,
num_classes,
is_bbox_normalized=is_bbox_normalized,
map_type=map_type)
box_ap_stats.append(box_ap)
return box_ap_stats