remove fields config in input_def (#1921)

* remove fields config in input_def

configs/_base_/datasets/coco.yml → configs/_base_/datasets/coco_detection.yml

@@ -6,6 +6,7 @@ TrainDataset:
     image_dir: train2017
     anno_path: annotations/instances_train2017.json
     dataset_dir: dataset/coco
+    data_fields: ['image', 'gt_bbox', 'gt_class', 'is_crowd']
 
 EvalDataset:
   !COCODataSet

configs/_base_/datasets/coco_instance.yml

+metric: COCO
+num_classes: 80
+
+TrainDataset:
+  !COCODataSet
+    image_dir: train2017
+    anno_path: annotations/instances_train2017.json
+    dataset_dir: dataset/coco
+    data_fields: ['image', 'gt_bbox', 'gt_class', 'gt_poly', 'is_crowd']
+
+EvalDataset:
+  !COCODataSet
+    image_dir: val2017
+    anno_path: annotations/instances_val2017.json
+    dataset_dir: dataset/coco
+
+TestDataset:
+  !ImageFolder
+    anno_path: annotations/instances_val2017.json

configs/_base_/datasets/voc.yml

@@ -6,12 +6,14 @@ TrainDataset:
     dataset_dir: dataset/voc
     anno_path: trainval.txt
     label_list: label_list.txt
+    data_fields: ['image', 'gt_bbox', 'gt_class', 'difficult']
 
 EvalDataset:
   !VOCDataSet
     dataset_dir: dataset/voc
     anno_path: test.txt
     label_list: label_list.txt
+    data_fields: ['image', 'gt_bbox', 'gt_class', 'difficult']
 
 TestDataset:
   !ImageFolder

configs/_base_/optimizers/ssd_240e.yml

+epoch: 240
+
+LearningRate:
+  base_lr: 0.001
+  schedulers:
+  - !PiecewiseDecay
+    gamma: 0.1
+    milestones:
+    - 160
+    - 200
+  - !LinearWarmup
+    start_factor: 0.3333333333333333
+    steps: 500
+
+OptimizerBuilder:
+  optimizer:
+    momentum: 0.9
+    type: Momentum
+  regularizer:
+    factor: 0.0005
+    type: L2

configs/_base_/readers/faster_fpn_reader.yml

 worker_num: 2
 TrainReader:
-  inputs_def:
-    fields: ['image', 'im_info', 'im_id', 'gt_bbox', 'gt_class', 'is_crowd']
   sample_transforms:
-  - DecodeImage: {to_rgb: true}
+  - DecodeOp: { }
   - RandomFlipImage: {prob: 0.5}
   - NormalizeImage: {is_channel_first: false, is_scale: true, mean: [0.485,0.456,0.406], std: [0.229, 0.224,0.225]}
   - ResizeImage: {target_size: 800, max_size: 1333, interp: 1, use_cv2: true}
@@ -16,8 +14,6 @@ TrainReader:
 
 
 EvalReader:
-  inputs_def:
-    fields: ['image', 'im_shape', 'scale_factor', 'im_id']
   sample_transforms:
   - DecodeOp: { }
   - NormalizeImageOp: { is_scale: true, mean: [ 0.485,0.456,0.406 ], std: [ 0.229, 0.224,0.225 ] }
@@ -32,8 +28,6 @@ EvalReader:
 
 
 TestReader:
-  inputs_def:
-    fields:  ['image', 'im_shape', 'scale_factor', 'im_id']
   sample_transforms:
   - DecodeOp: { }
   - NormalizeImageOp: { is_scale: true, mean: [ 0.485,0.456,0.406 ], std: [ 0.229, 0.224,0.225 ] }

configs/_base_/readers/faster_reader.yml

 worker_num: 2
 TrainReader:
-  inputs_def:
-    fields: ['image', 'im_info', 'im_id', 'gt_bbox', 'gt_class', 'is_crowd']
   sample_transforms:
-  - DecodeImage: {to_rgb: true}
+  - DecodeOp: { }
   - RandomFlipImage: {prob: 0.5}
   - NormalizeImage: {is_channel_first: false, is_scale: true, mean: [0.485,0.456,0.406], std: [0.229, 0.224,0.225]}
   - ResizeImage: {target_size: 800, max_size: 1333, interp: 1, use_cv2: true}
@@ -16,8 +14,6 @@ TrainReader:
 
 
 EvalReader:
-  inputs_def:
-    fields: ['image', 'im_shape', 'scale_factor', 'im_id']
   sample_transforms:
   - DecodeOp: { }
   - NormalizeImageOp: { is_scale: true, mean: [ 0.485,0.456,0.406 ], std: [ 0.229, 0.224,0.225 ] }
@@ -32,8 +28,6 @@ EvalReader:
 
 
 TestReader:
-  inputs_def:
-    fields: ['image', 'im_shape', 'scale_factor', 'im_id']
   sample_transforms:
   - DecodeOp: { }
   - NormalizeImageOp: { is_scale: true, mean: [ 0.485,0.456,0.406 ], std: [ 0.229, 0.224,0.225 ] }

configs/_base_/readers/mask_fpn_reader.yml

 worker_num: 2
 TrainReader:
-  inputs_def:
-    fields: ['image', 'im_info', 'im_id', 'gt_bbox', 'gt_class', 'is_crowd', 'gt_poly']
   sample_transforms:
-  - DecodeImage: {to_rgb: true}
+  - DecodeOp: {}
   - RandomFlipImage: {prob: 0.5, is_mask_flip: true}
   - NormalizeImage: {is_channel_first: false, is_scale: true, mean: [0.485,0.456,0.406], std: [0.229, 0.224,0.225]}
   - ResizeImage: {target_size: 800, max_size: 1333, interp: 1, use_cv2: true}
@@ -16,8 +14,6 @@ TrainReader:
 
 
 EvalReader:
-  inputs_def:
-    fields: ['image', 'im_shape', 'scale_factor', 'im_id']
   sample_transforms:
   - DecodeOp: {}
   - NormalizeImageOp: {is_scale: true, mean: [0.485,0.456,0.406], std: [0.229, 0.224,0.225]}
@@ -32,8 +28,6 @@ EvalReader:
 
 
 TestReader:
-  inputs_def:
-    fields: ['image', 'im_shape', 'scale_factor', 'im_id']
   sample_transforms:
   - DecodeOp: {}
   - NormalizeImageOp: {is_scale: true, mean: [0.485,0.456,0.406], std: [0.229, 0.224,0.225]}

configs/_base_/readers/mask_reader.yml

 worker_num: 2
 TrainReader:
-  inputs_def:
-    fields: ['image', 'im_info', 'im_id', 'gt_bbox', 'gt_class', 'is_crowd', 'gt_poly']
   sample_transforms:
-  - DecodeImage: {to_rgb: true}
+  - DecodeOp: {}
   - RandomFlipImage: {prob: 0.5, is_mask_flip: true}
   - NormalizeImage: {is_channel_first: false, is_scale: true, mean: [0.485,0.456,0.406], std: [0.229, 0.224,0.225]}
   - ResizeImage: {target_size: 800, max_size: 1333, interp: 1, use_cv2: true}
@@ -16,8 +14,6 @@ TrainReader:
 
 
 EvalReader:
-  inputs_def:
-    fields: ['image', 'im_shape', 'scale_factor', 'im_id']
   sample_transforms:
   - DecodeOp: {}
   - NormalizeImageOp: {is_scale: true, mean: [0.485,0.456,0.406], std: [0.229, 0.224,0.225]}
@@ -32,8 +28,6 @@ EvalReader:
 
 
 TestReader:
-  inputs_def:
-    fields: ['image', 'im_shape', 'scale_factor', 'im_id']
   sample_transforms:
   - DecodeOp: {}
   - NormalizeImageOp: {is_scale: true, mean: [0.485,0.456,0.406], std: [0.229, 0.224,0.225]}

configs/_base_/readers/ssd_reader.yml

 worker_num: 2
 TrainReader:
   inputs_def:
-    fields: ['image', 'gt_bbox', 'gt_class']
     num_max_boxes: 90
 
   sample_transforms:
@@ -24,8 +23,6 @@ TrainReader:
 
 
 EvalReader:
-  inputs_def:
-    fields: ['image', 'im_shape',  'scale_factor', 'im_id', 'gt_bbox', 'gt_class', 'difficult']
   sample_transforms:
     - DecodeOp: {}
     - ResizeOp: {target_size: [300, 300], keep_ratio: False, interp: 1}
@@ -37,7 +34,6 @@ EvalReader:
 TestReader:
   inputs_def:
     image_shape: [3, 300, 300]
-    fields: ['image', 'im_shape', 'scale_factor', 'im_id']
   sample_transforms:
     - DecodeOp: {}
     - ResizeOp: {target_size: [300, 300], keep_ratio: False, interp: 1}

configs/_base_/readers/yolov3_reader.yml

 worker_num: 2
 TrainReader:
   inputs_def:
-    fields: ['image', 'gt_bbox', 'gt_class', 'gt_score', 'im_shape', 'scale_factor']
     num_max_boxes: 50
   sample_transforms:
     - DecodeOp: {}
@@ -26,13 +25,11 @@ TrainReader:
 
 EvalReader:
   inputs_def:
-    fields: ['image', 'im_shape', 'scale_factor', 'im_id']
     num_max_boxes: 50
   sample_transforms:
     - DecodeOp: {}
     - ResizeOp: {target_size: [608, 608], keep_ratio: False, interp: 2}
     - NormalizeImageOp: {mean: [0.485, 0.456, 0.406], std: [0.229, 0.224, 0.225], is_scale: True}
-    - PadBoxOp: {num_max_boxes: 50}
     - PermuteOp: {}
   batch_size: 1
   drop_empty: false
@@ -40,7 +37,6 @@ EvalReader:
 TestReader:
   inputs_def:
     image_shape: [3, 608, 608]
-    fields: ['image', 'im_shape', 'scale_factor', 'im_id']
   sample_transforms:
     - DecodeOp: {}
     - ResizeOp: {target_size: [608, 608], keep_ratio: False, interp: 2}

configs/cascade_mask_rcnn_r50_fpn_1x_coco.yml

 _BASE_: [
   './_base_/models/cascade_mask_rcnn_r50_fpn.yml',
   './_base_/optimizers/rcnn_1x.yml',
-  './_base_/datasets/coco.yml',
+  './_base_/datasets/coco_instance.yml',
   './_base_/readers/mask_fpn_reader.yml',
   './_base_/runtime.yml',
 ]

configs/cascade_rcnn_r50_fpn_1x_coco.yml

 _BASE_: [
   './_base_/models/cascade_rcnn_r50_fpn.yml',
   './_base_/optimizers/rcnn_1x.yml',
-  './_base_/datasets/coco.yml',
+  './_base_/datasets/coco_detection.yml',
   './_base_/readers/faster_fpn_reader.yml',
   './_base_/runtime.yml',
 ]

configs/faster_rcnn_r50_1x_coco.yml

 _BASE_: [
   './_base_/models/faster_rcnn_r50.yml',
   './_base_/optimizers/rcnn_1x.yml',
-  './_base_/datasets/coco.yml',
+  './_base_/datasets/coco_detection.yml',
   './_base_/readers/faster_reader.yml',
   './_base_/runtime.yml',
 ]

configs/faster_rcnn_r50_fpn_1x_coco.yml

 _BASE_: [
   './_base_/models/faster_rcnn_r50_fpn.yml',
   './_base_/optimizers/rcnn_1x.yml',
-  './_base_/datasets/coco.yml',
+  './_base_/datasets/coco_detection.yml',
   './_base_/readers/faster_fpn_reader.yml',
   './_base_/runtime.yml',
 ]

configs/mask_rcnn_r50_1x_coco.yml

 _BASE_: [
   './_base_/models/mask_rcnn_r50.yml',
   './_base_/optimizers/rcnn_1x.yml',
-  './_base_/datasets/coco.yml',
+  './_base_/datasets/coco_instance.yml',
   './_base_/readers/mask_reader.yml',
   './_base_/runtime.yml',
 ]

configs/mask_rcnn_r50_fpn_1x_coco.yml

 _BASE_: [
   './_base_/models/mask_rcnn_r50_fpn.yml',
   './_base_/optimizers/rcnn_1x.yml',
-  './_base_/datasets/coco.yml',
+  './_base_/datasets/coco_instance.yml',
   './_base_/readers/mask_fpn_reader.yml',
   './_base_/runtime.yml',
 ]

configs/ssd_vgg16_300_120e_coco.yml

-_BASE_: [
-  './_base_/models/ssd_vgg16_300.yml',
-  './_base_/optimizers/ssd_120e.yml',
-  './_base_/datasets/coco.yml',
-  './_base_/readers/ssd_reader.yml',
-  './_base_/runtime.yml',
-]

configs/yolov3_darknet53_270e_coco.yml

 _BASE_: [
   './_base_/models/yolov3_darknet53.yml',
   './_base_/optimizers/yolov3_270e.yml',
-  './_base_/datasets/coco.yml',
+  './_base_/datasets/coco_detection.yml',
   './_base_/readers/yolov3_reader.yml',
   './_base_/runtime.yml',
 ]

ppdet/data/reader.py

@@ -16,6 +16,7 @@ import copy
 import traceback
 import six
 import sys
+import multiprocessing as mp
 if sys.version_info >= (3, 0):
     import queue as Queue
 else:
@@ -27,45 +28,42 @@ from paddle.io import DistributedBatchSampler
 
 from ppdet.core.workspace import register, serializable, create
 from . import transform
-from .transform import operator, batch_operator
 
 from ppdet.utils.logger import setup_logger
 logger = setup_logger('reader')
 
 
 class Compose(object):
-    def __init__(self, transforms, fields=None, from_=transform,
-                 num_classes=81):
+    def __init__(self, transforms, num_classes=81):
         self.transforms = transforms
         self.transforms_cls = []
-        output_fields = None
         for t in self.transforms:
             for k, v in t.items():
-                op_cls = getattr(from_, k)
+                op_cls = getattr(transform, k)
                 self.transforms_cls.append(op_cls(**v))
                 if hasattr(op_cls, 'num_classes'):
                     op_cls.num_classes = num_classes
 
-                # TODO: should be refined in the future
-                if op_cls in [
-                        transform.Gt2YoloTargetOp, transform.Gt2YoloTarget
-                ]:
-                    output_fields = ['image', 'gt_bbox']
-                    output_fields.extend([
-                        'target{}'.format(i)
-                        for i in range(len(v['anchor_masks']))
-                    ])
+    def __call__(self, data):
+        for f in self.transforms_cls:
+            try:
+                data = f(data)
+            except Exception as e:
+                stack_info = traceback.format_exc()
+                logger.warn("fail to map op [{}] with error: {} and stack:\n{}".
+                            format(f, e, str(stack_info)))
+                raise e
+
+        return data
 
-        self.fields = fields
-        self.output_fields = output_fields if output_fields else fields
 
-    def __call__(self, data):
-        if self.fields is not None:
-            data_new = []
-            for item in data:
-                data_new.append(dict(zip(self.fields, item)))
-            data = data_new
+class BatchCompose(Compose):
+    def __init__(self, transforms, num_classes=81):
+        super(BatchCompose, self).__init__(transforms, num_classes)
+        self.output_fields = mp.Manager().list([])
+        self.lock = mp.Lock()
 
+    def __call__(self, data):
         for f in self.transforms_cls:
             try:
                 data = f(data)
@@ -75,23 +73,27 @@ class Compose(object):
                             format(f, e, str(stack_info)))
                 raise e
 
-        if self.output_fields is not None:
-            data_new = []
-            for item in data:
-                batch = []
-                for k in self.output_fields:
-                    batch.append(item[k])
-                data_new.append(batch)
-            batch_size = len(data_new)
-            data_new = list(zip(*data_new))
-            if batch_size > 1:
-                data = [
-                    np.array(item).astype(item[0].dtype) for item in data_new
-                ]
-            else:
-                data = data_new
-
-        return data
+        # parse output fields by first sample
+        # **this shoule be fixed if paddle.io.DataLoader support**
+        # For paddle.io.DataLoader not support dict currently,
+        # we need to parse the key from the first sample,
+        # BatchCompose.__call__ will be called in each worker
+        # process, so lock is need here.
+        if len(self.output_fields) == 0:
+            self.lock.acquire()
+            if len(self.output_fields) == 0:
+                for k, v in data[0].items():
+                    # FIXME(dkp): for more elegent coding
+                    if k not in ['flipped', 'h', 'w']:
+                        self.output_fields.append(k)
+            self.lock.release()
+
+        data = [[data[i][k] for k in self.output_fields]
+                for i in range(len(data))]
+        data = list(zip(*data))
+
+        batch_data = [np.stack(d, axis=0) for d in data]
+        return batch_data
 
 
 class BaseDataLoader(object):
@@ -99,8 +101,8 @@ class BaseDataLoader(object):
 
     def __init__(self,
                  inputs_def=None,
-                 sample_transforms=None,
-                 batch_transforms=None,
+                 sample_transforms=[],
+                 batch_transforms=[],
                  batch_size=1,
                  shuffle=False,
                  drop_last=False,
@@ -108,21 +110,12 @@ class BaseDataLoader(object):
                  num_classes=81,
                  with_background=True,
                  **kwargs):
-        # out fields 
-        self._fields = inputs_def['fields'] if inputs_def else None
         # sample transform
         self._sample_transforms = Compose(
             sample_transforms, num_classes=num_classes)
 
         # batch transfrom 
-        self._batch_transforms = None
-        if batch_transforms:
-            self._batch_transforms = Compose(batch_transforms,
-                                             copy.deepcopy(self._fields),
-                                             transform, num_classes)
-            self.output_fields = self._batch_transforms.output_fields
-        else:
-            self.output_fields = self._fields
+        self._batch_transforms = BatchCompose(batch_transforms, num_classes)
 
         self.batch_size = batch_size
         self.shuffle = shuffle
@@ -139,8 +132,7 @@ class BaseDataLoader(object):
         self.dataset = dataset
         self.dataset.parse_dataset(self.with_background)
         # get data
-        self.dataset.set_out(self._sample_transforms,
-                             copy.deepcopy(self._fields))
+        self.dataset.set_transform(self._sample_transforms)
         # set kwargs
         self.dataset.set_kwargs(**self.kwargs)
         # batch sampler
@@ -177,7 +169,10 @@ class BaseDataLoader(object):
         # data structure in paddle.io.DataLoader
         try:
             data = next(self.loader)
-            return {k: v for k, v in zip(self.output_fields, data)}
+            return {
+                k: v
+                for k, v in zip(self._batch_transforms.output_fields, data)
+            }
         except StopIteration:
             self.loader = iter(self.dataloader)
             six.reraise(*sys.exc_info())
@@ -191,8 +186,8 @@ class BaseDataLoader(object):
 class TrainReader(BaseDataLoader):
     def __init__(self,
                  inputs_def=None,
-                 sample_transforms=None,
-                 batch_transforms=None,
+                 sample_transforms=[],
+                 batch_transforms=[],
                  batch_size=1,
                  shuffle=True,
                  drop_last=True,
@@ -210,8 +205,8 @@ class TrainReader(BaseDataLoader):
 class EvalReader(BaseDataLoader):
     def __init__(self,
                  inputs_def=None,
-                 sample_transforms=None,
-                 batch_transforms=None,
+                 sample_transforms=[],
+                 batch_transforms=[],
                  batch_size=1,
                  shuffle=False,
                  drop_last=True,
@@ -229,8 +224,8 @@ class EvalReader(BaseDataLoader):
 class TestReader(BaseDataLoader):
     def __init__(self,
                  inputs_def=None,
-                 sample_transforms=None,
-                 batch_transforms=None,
+                 sample_transforms=[],
+                 batch_transforms=[],
                  batch_size=1,
                  shuffle=False,
                  drop_last=False,

ppdet/data/source/coco.py

@@ -28,9 +28,10 @@ class COCODataSet(DetDataset):
                  dataset_dir=None,
                  image_dir=None,
                  anno_path=None,
+                 data_fields=['image'],
                  sample_num=-1):
         super(COCODataSet, self).__init__(dataset_dir, image_dir, anno_path,
-                                          sample_num)
+                                          data_fields, sample_num)
         self.load_image_only = False
         self.load_semantic = False
 
@@ -82,13 +83,6 @@ class COCODataSet(DetDataset):
                                    im_w, im_h, img_id))
                 continue
 
-            coco_rec = {
-                'im_file': im_path,
-                'im_id': np.array([img_id]),
-                'h': im_h,
-                'w': im_w,
-            }
-
             if not self.load_image_only:
                 ins_anno_ids = coco.getAnnIds(imgIds=img_id, iscrowd=False)
                 instances = coco.loadAnns(ins_anno_ids)
@@ -121,7 +115,6 @@ class COCODataSet(DetDataset):
 
                 gt_bbox = np.zeros((num_bbox, 4), dtype=np.float32)
                 gt_class = np.zeros((num_bbox, 1), dtype=np.int32)
-                gt_score = np.ones((num_bbox, 1), dtype=np.float32)
                 is_crowd = np.zeros((num_bbox, 1), dtype=np.int32)
                 difficult = np.zeros((num_bbox, 1), dtype=np.int32)
                 gt_poly = [None] * num_bbox
@@ -142,15 +135,25 @@ class COCODataSet(DetDataset):
                 if has_segmentation and not any(gt_poly):
                     continue
 
-                coco_rec.update({
+                coco_rec = {
+                    'im_file': im_path,
+                    'im_id': np.array([img_id]),
+                    'h': im_h,
+                    'w': im_w,
+                } if 'image' in self.data_fields else {}
+
+                gt_rec = {
                     'is_crowd': is_crowd,
                     'gt_class': gt_class,
                     'gt_bbox': gt_bbox,
-                    'gt_score': gt_score,
                     'gt_poly': gt_poly,
-                })
+                }
+                for k, v in gt_rec.items():
+                    if k in self.data_fields:
+                        coco_rec[k] = v
+
                 # TODO: remove load_semantic
-                if self.load_semantic:
+                if self.load_semantic and 'semantic' in self.data_fields:
                     seg_path = os.path.join(self.dataset_dir, 'stuffthingmaps',
                                             'train2017', im_fname[:-3] + 'png')
                     coco_rec.update({'semantic': seg_path})

ppdet/data/source/dataset.py

@@ -31,6 +31,7 @@ class DetDataset(Dataset):
                  dataset_dir=None,
                  image_dir=None,
                  anno_path=None,
+                 data_fields=['image'],
                  sample_num=-1,
                  use_default_label=None,
                  **kwargs):
@@ -38,6 +39,7 @@ class DetDataset(Dataset):
         self.dataset_dir = dataset_dir if dataset_dir is not None else ''
         self.anno_path = anno_path
         self.image_dir = image_dir if image_dir is not None else ''
+        self.data_fields = data_fields
         self.sample_num = sample_num
         self.use_default_label = use_default_label
         self._epoch = 0
@@ -63,26 +65,19 @@ class DetDataset(Dataset):
                 for _ in range(3)
             ]
 
-        # data augment
-        roidb = self.transform(roidb)
-        # data item 
-        out = OrderedDict()
-        for k in self.fields:
-            out[k] = roidb[k]
-        return out.values()
+        return self.transform(roidb)
 
     def set_kwargs(self, **kwargs):
         self.mixup_epoch = kwargs.get('mixup_epoch', -1)
         self.cutmix_epoch = kwargs.get('cutmix_epoch', -1)
         self.mosaic_epoch = kwargs.get('mosaic_epoch', -1)
 
+    def set_transform(self, transform):
+        self.transform = transform
+
     def set_epoch(self, epoch_id):
         self._epoch = epoch_id
 
-    def set_out(self, sample_transform, fields):
-        self.transform = sample_transform
-        self.fields = fields
-
     def parse_dataset(self, with_background=True):
         raise NotImplemented(
             "Need to implement parse_dataset method of Dataset")

ppdet/data/source/voc.py

@@ -46,12 +46,14 @@ class VOCDataSet(DetDataset):
                  dataset_dir=None,
                  image_dir=None,
                  anno_path=None,
+                 data_fields=['image'],
                  sample_num=-1,
                  label_list=None):
         super(VOCDataSet, self).__init__(
             dataset_dir=dataset_dir,
             image_dir=image_dir,
             anno_path=anno_path,
+            data_fields=data_fields,
             sample_num=sample_num)
         self.label_list = label_list
 
@@ -113,7 +115,6 @@ class VOCDataSet(DetDataset):
                 gt_bbox = []
                 gt_class = []
                 gt_score = []
-                is_crowd = []
                 difficult = []
                 for i, obj in enumerate(objs):
                     cname = obj.find('name').text
@@ -130,7 +131,6 @@ class VOCDataSet(DetDataset):
                         gt_bbox.append([x1, y1, x2, y2])
                         gt_class.append([cname2cid[cname]])
                         gt_score.append([1.])
-                        is_crowd.append([0])
                         difficult.append([_difficult])
                     else:
                         logger.warn(
@@ -140,19 +140,25 @@ class VOCDataSet(DetDataset):
                 gt_bbox = np.array(gt_bbox).astype('float32')
                 gt_class = np.array(gt_class).astype('int32')
                 gt_score = np.array(gt_score).astype('float32')
-                is_crowd = np.array(is_crowd).astype('int32')
                 difficult = np.array(difficult).astype('int32')
+
                 voc_rec = {
                     'im_file': img_file,
                     'im_id': im_id,
                     'h': im_h,
-                    'w': im_w,
-                    'is_crowd': is_crowd,
+                    'w': im_w
+                } if 'image' in self.data_fields else {}
+
+                gt_rec = {
                     'gt_class': gt_class,
                     'gt_score': gt_score,
                     'gt_bbox': gt_bbox,
                     'difficult': difficult
                 }
+                for k, v in gt_rec.items():
+                    if k in self.data_fields:
+                        voc_rec[k] = v
+
                 if len(objs) != 0:
                     records.append(voc_rec)
 

ppdet/data/transform/batch_operator.py

@@ -303,6 +303,11 @@ class Gt2YoloTargetOp(BaseOperator):
                                 # classification
                                 target[idx, 6 + cls, gj, gi] = 1.
                 sample['target{}'.format(i)] = target
+
+            # remove useless gt_class and gt_score after target calculated
+            sample.pop('gt_class')
+            sample.pop('gt_score')
+
         return samples
 
 

ppdet/data/transform/operator.py

@@ -116,6 +116,7 @@ class DecodeOp(BaseOperator):
         if 'image' not in sample:
             with open(sample['im_file'], 'rb') as f:
                 sample['image'] = f.read()
+            sample.pop('im_file')
 
         im = sample['image']
         data = np.frombuffer(im, dtype='uint8')
@@ -1570,9 +1571,9 @@ class MixupOp(BaseOperator):
             gt_class2 = sample[1]['gt_class']
             gt_class = np.concatenate((gt_class1, gt_class2), axis=0)
             result['gt_class'] = gt_class
-        if 'gt_score' in sample[0]:
-            gt_score1 = sample[0]['gt_score']
-            gt_score2 = sample[1]['gt_score']
+
+            gt_score1 = np.ones_like(sample[0]['gt_class'])
+            gt_score2 = np.ones_like(sample[1]['gt_class'])
             gt_score = np.concatenate(
                 (gt_score1 * factor, gt_score2 * (1. - factor)), axis=0)
             result['gt_score'] = gt_score
@@ -1673,6 +1674,11 @@ class PadBoxOp(BaseOperator):
             if gt_num > 0:
                 pad_diff[:gt_num] = sample['difficult'][:gt_num, 0]
             sample['difficult'] = pad_diff
+        if 'is_crowd' in sample:
+            pad_crowd = np.zeros((num_max, ), dtype=np.int32)
+            if gt_num > 0:
+                pad_crowd[:gt_num] = sample['is_crowd'][:gt_num, 0]
+            sample['is_crowd'] = pad_crowd
         return sample
 
 

tools/infer.py

@@ -131,16 +131,18 @@ def run(FLAGS, cfg, place):
     dataset.set_images(test_images)
     test_loader = create('TestReader')(dataset, cfg['worker_num'])
     extra_key = ['im_shape', 'scale_factor', 'im_id']
-    if cfg.metric == 'VOC':
-        extra_key += ['gt_bbox', 'gt_class', 'difficult']
 
     # TODO: support other metrics
     imid2path = dataset.get_imid2path()
 
-    from ppdet.utils.coco_eval import get_category_info
     anno_file = dataset.get_anno()
     with_background = cfg.with_background
     use_default_label = dataset.use_default_label
+
+    if cfg.metric == 'COCO':
+        from ppdet.utils.coco_eval import get_category_info
+    if cfg.metric == 'VOC':
+        from ppdet.utils.voc_eval import get_category_info
     clsid2catid, catid2name = get_category_info(anno_file, with_background,
                                                 use_default_label)