def create_model(weights):
backbone = build_backbone()
transformer = Transformer(d_model=256, return_intermediate_dec=True)
model = DETR(backbone, transformer, num_classes=91, num_queries=100)
checkpoint = torch.load(weights, map_location='cpu')['model']
model.load_state_dict(checkpoint)
return model
def main():
num_classes = 91
device = torch.device('cuda')
backbone = build_backbone()
transformer = Transformer(
d_model=256,
dropout=0.1,
nhead=8,
dim_feedforward=2048,
num_encoder_layers=6,
num_decoder_layers=6,
normalize_before=False,
return_intermediate_dec=True,
)
model = DETR(
backbone,
transformer,
num_classes=num_classes,
num_queries=100,
aux_loss=True,
)
checkpoint = torch.load('./detr-r50-e632da11.pth')
model.load_state_dict(checkpoint['model'])
model.to(device)
model.eval()
gen_wts(model, "detr")
def build_paddle_model():
# * backbone
backbone = 'resnet50'
dilation = False
position_embedding = 'sine'
# * Transformer
enc_layers = 6
dec_layers = 6
dim_feedforward = 2048
hidden_dim = 256
dropout = 0
nheads = 8
num_queries = 100
pre_norm = False
num_classes = 91
position_embedding = PositionEmbeddingSine(hidden_dim // 2, normalize=True)
backbone = Backbone(backbone, False, True, dilation)
backbone = Joiner(backbone, position_embedding)
backbone.num_channels = 2048
transformer = Transformer(
d_model=hidden_dim, dropout=dropout, nhead=nheads, dim_feedforward=dim_feedforward,
num_encoder_layers=enc_layers, num_decoder_layers=dec_layers, normalize_before=pre_norm,
return_intermediate_dec=True)
model = DETR(backbone, transformer, num_classes=num_classes, num_queries=num_queries, aux_loss=True)
return model
def _make_detr(backbone_name: str, dilation=False, num_classes=91, mask=False):
hidden_dim = 256
backbone = Backbone(backbone_name, train_backbone=True, return_interm_layers=mask, dilation=dilation)
pos_enc = PositionEmbeddingSine(hidden_dim // 2, normalize=True)
backbone_with_pos_enc = Joiner(backbone, pos_enc)
backbone_with_pos_enc.num_channels = backbone.num_channels
transformer = Transformer(d_model=hidden_dim, return_intermediate_dec=True)
detr = DETR(backbone_with_pos_enc, transformer, num_classes=num_classes, num_queries=100)
if mask:
return DETRsegm(detr)
return detr
dropout = 0.1
nheads = 8
dim_feedforward = 2048
enc_layers = 6
dec_layers = 6
pre_norm = False
num_classes = 3
num_queries = 100
aux_loss = True
backbone = build_backbone(lr_backbone, masks, backbone, dilation, hidden_dim, position_embedding)
transformer = build_transformer(hidden_dim, dropout, nheads, dim_feedforward, enc_layers, dec_layers, pre_norm)
model = DETR(
backbone,
transformer,
num_classes=num_classes,
num_queries=num_queries,
aux_loss=aux_loss,
)
transform = make_coco_transforms()
postprocessors = PostProcess()
checkpoint = torch.load('/home/palm/PycharmProjects/detr/snapshots/1/checkpoint00295.pth')
model.load_state_dict(checkpoint['model'])
train_ints, valid_ints, labels, max_box_per_image = create_csv_training_instances(
'/home/palm/PycharmProjects/algea/dataset/train_annotations',
'/home/palm/PycharmProjects/algea/dataset/test_annotations',
'/home/palm/PycharmProjects/algea/dataset/classes',
)
# os.listdir()
def __init__(self, cfg):
super().__init__()
self.device = torch.device(cfg.MODEL.DEVICE)
self.num_classes = cfg.MODEL.DETR.NUM_CLASSES
hidden_dim = cfg.MODEL.DETR.HIDDEN_DIM
num_queries = cfg.MODEL.DETR.NUM_OBJECT_QUERIES
# Transformer parameters:
nheads = cfg.MODEL.DETR.NHEADS
dropout = cfg.MODEL.DETR.DROPOUT
dim_feedforward = cfg.MODEL.DETR.DIM_FEEDFORWARD
enc_layers = cfg.MODEL.DETR.ENC_LAYERS
dec_layers = cfg.MODEL.DETR.DEC_LAYERS
pre_norm = cfg.MODEL.DETR.PRE_NORM
pass_pos_and_query = cfg.MODEL.DETR.PASS_POS_AND_QUERY
# Loss parameters:
giou_weight = cfg.MODEL.DETR.GIOU_WEIGHT
l1_weight = cfg.MODEL.DETR.L1_WEIGHT
deep_supervision = cfg.MODEL.DETR.DEEP_SUPERVISION
no_object_weight = cfg.MODEL.DETR.NO_OBJECT_WEIGHT
N_steps = hidden_dim // 2
d2_backbone = MaskedBackbone(cfg)
backbone = Joiner(d2_backbone,
PositionEmbeddingSine(N_steps, normalize=True))
backbone.num_channels = d2_backbone.num_channels
transformer = Transformer(
d_model=hidden_dim,
dropout=dropout,
nhead=nheads,
dim_feedforward=dim_feedforward,
num_encoder_layers=enc_layers,
num_decoder_layers=dec_layers,
normalize_before=pre_norm,
return_intermediate_dec=deep_supervision,
pass_pos_and_query=pass_pos_and_query,
)
self.detr = DETR(backbone,
transformer,
num_classes=self.num_classes,
num_queries=num_queries,
aux_loss=deep_supervision)
self.detr.to(self.device)
# building criterion
matcher = HungarianMatcher(cost_class=1,
cost_bbox=l1_weight,
cost_giou=giou_weight)
weight_dict = {"loss_ce": 1, "loss_bbox": l1_weight}
weight_dict["loss_giou"] = giou_weight
if deep_supervision:
aux_weight_dict = {}
for i in range(dec_layers - 1):
aux_weight_dict.update(
{k + f"_{i}": v
for k, v in weight_dict.items()})
weight_dict.update(aux_weight_dict)
losses = ["labels", "boxes", "cardinality"]
self.criterion = SetCriterion(self.num_classes,
matcher=matcher,
weight_dict=weight_dict,
eos_coef=no_object_weight,
losses=losses)
self.criterion.to(self.device)
pixel_mean = torch.Tensor(cfg.MODEL.PIXEL_MEAN).to(self.device).view(
3, 1, 1)
pixel_std = torch.Tensor(cfg.MODEL.PIXEL_STD).to(self.device).view(
3, 1, 1)
self.normalizer = lambda x: (x - pixel_mean) / pixel_std
self.to(self.device)
class Detr(nn.Module):
"""
Implement Detr
"""
def __init__(self, cfg):
super().__init__()
self.device = torch.device(cfg.MODEL.DEVICE)
self.num_classes = cfg.MODEL.DETR.NUM_CLASSES
hidden_dim = cfg.MODEL.DETR.HIDDEN_DIM
num_queries = cfg.MODEL.DETR.NUM_OBJECT_QUERIES
# Transformer parameters:
nheads = cfg.MODEL.DETR.NHEADS
dropout = cfg.MODEL.DETR.DROPOUT
dim_feedforward = cfg.MODEL.DETR.DIM_FEEDFORWARD
enc_layers = cfg.MODEL.DETR.ENC_LAYERS
dec_layers = cfg.MODEL.DETR.DEC_LAYERS
pre_norm = cfg.MODEL.DETR.PRE_NORM
pass_pos_and_query = cfg.MODEL.DETR.PASS_POS_AND_QUERY
# Loss parameters:
giou_weight = cfg.MODEL.DETR.GIOU_WEIGHT
l1_weight = cfg.MODEL.DETR.L1_WEIGHT
deep_supervision = cfg.MODEL.DETR.DEEP_SUPERVISION
no_object_weight = cfg.MODEL.DETR.NO_OBJECT_WEIGHT
N_steps = hidden_dim // 2
d2_backbone = MaskedBackbone(cfg)
backbone = Joiner(d2_backbone,
PositionEmbeddingSine(N_steps, normalize=True))
backbone.num_channels = d2_backbone.num_channels
transformer = Transformer(
d_model=hidden_dim,
dropout=dropout,
nhead=nheads,
dim_feedforward=dim_feedforward,
num_encoder_layers=enc_layers,
num_decoder_layers=dec_layers,
normalize_before=pre_norm,
return_intermediate_dec=deep_supervision,
pass_pos_and_query=pass_pos_and_query,
)
self.detr = DETR(backbone,
transformer,
num_classes=self.num_classes,
num_queries=num_queries,
aux_loss=deep_supervision)
self.detr.to(self.device)
# building criterion
matcher = HungarianMatcher(cost_class=1,
cost_bbox=l1_weight,
cost_giou=giou_weight)
weight_dict = {"loss_ce": 1, "loss_bbox": l1_weight}
weight_dict["loss_giou"] = giou_weight
if deep_supervision:
aux_weight_dict = {}
for i in range(dec_layers - 1):
aux_weight_dict.update(
{k + f"_{i}": v
for k, v in weight_dict.items()})
weight_dict.update(aux_weight_dict)
losses = ["labels", "boxes", "cardinality"]
self.criterion = SetCriterion(self.num_classes,
matcher=matcher,
weight_dict=weight_dict,
eos_coef=no_object_weight,
losses=losses)
self.criterion.to(self.device)
pixel_mean = torch.Tensor(cfg.MODEL.PIXEL_MEAN).to(self.device).view(
3, 1, 1)
pixel_std = torch.Tensor(cfg.MODEL.PIXEL_STD).to(self.device).view(
3, 1, 1)
self.normalizer = lambda x: (x - pixel_mean) / pixel_std
self.to(self.device)
def forward(self, batched_inputs):
"""
Args:
batched_inputs: a list, batched outputs of :class:`DatasetMapper` .
Each item in the list contains the inputs for one image.
For now, each item in the list is a dict that contains:
* image: Tensor, image in (C, H, W) format.
* instances: Instances
Other information that's included in the original dicts, such as:
* "height", "width" (int): the output resolution of the model, used in inference.
See :meth:`postprocess` for details.
Returns:
dict[str: Tensor]:
mapping from a named loss to a tensor storing the loss. Used during training only.
"""
images = self.preprocess_image(batched_inputs)
output = self.detr(images)
if self.training:
gt_instances = [
x["instances"].to(self.device) for x in batched_inputs
]
targets = self.prepare_targets(gt_instances)
loss_dict = self.criterion(output, targets)
weight_dict = self.criterion.weight_dict
for k in loss_dict.keys():
if k in weight_dict:
loss_dict[k] *= weight_dict[k]
return loss_dict
else:
box_cls = output["pred_logits"]
box_pred = output["pred_boxes"]
results = self.inference(box_cls, box_pred, images.image_sizes)
processed_results = []
for results_per_image, input_per_image, image_size in zip(
results, batched_inputs, images.image_sizes):
height = input_per_image.get("height", image_size[0])
width = input_per_image.get("width", image_size[1])
r = detector_postprocess(results_per_image, height, width)
processed_results.append({"instances": r})
return processed_results
def prepare_targets(self, targets):
new_targets = []
for targets_per_image in targets:
h, w = targets_per_image.image_size
image_size_xyxy = torch.as_tensor([w, h, w, h],
dtype=torch.float,
device=self.device)
gt_classes = targets_per_image.gt_classes
gt_boxes = targets_per_image.gt_boxes.tensor / image_size_xyxy
gt_boxes = box_xyxy_to_cxcywh(gt_boxes)
new_targets.append({"labels": gt_classes, "boxes": gt_boxes})
return new_targets
def inference(self, box_cls, box_pred, image_sizes):
"""
Arguments:
box_cls (Tensor): tensor of shape (batch_size, num_queries, K).
The tensor predicts the classification probability for each query.
box_pred (Tensor): tensors of shape (batch_size, num_queries, 4).
The tensor predicts 4-vector (x,y,w,h) box
regression values for every queryx
image_sizes (List[torch.Size]): the input image sizes
Returns:
results (List[Instances]): a list of #images elements.
"""
assert len(box_cls) == len(image_sizes)
results = []
# For each box we assign the best class or the second best if the best on is `no_object`.
scores, labels = F.softmax(box_cls, dim=-1)[:, :, :-1].max(-1)
for scores_per_image, labels_per_image, box_pred_per_image, image_size in zip(
scores, labels, box_pred, image_sizes):
result = Instances(image_size)
result.pred_boxes = Boxes(box_cxcywh_to_xyxy(box_pred_per_image))
result.pred_boxes.scale(scale_x=image_size[1],
scale_y=image_size[0])
result.scores = scores_per_image
result.pred_classes = labels_per_image
results.append(result)
return results
def preprocess_image(self, batched_inputs):
"""
Normalize, pad and batch the input images.
"""
images = [
self.normalizer(x["image"].to(self.device)) for x in batched_inputs
]
images = ImageList.from_tensors(images)
return images
def __init__(self, cfg):
super().__init__()
self.device = torch.device(cfg.MODEL.DEVICE)
self.num_classes = cfg.MODEL.DETR.NUM_CLASSES
self.mask_on = cfg.MODEL.MASK_ON
hidden_dim = cfg.MODEL.DETR.HIDDEN_DIM
num_queries = cfg.MODEL.DETR.NUM_OBJECT_QUERIES
# Transformer parameters:
nheads = cfg.MODEL.DETR.NHEADS
dropout = cfg.MODEL.DETR.DROPOUT
dim_feedforward = cfg.MODEL.DETR.DIM_FEEDFORWARD
enc_layers = cfg.MODEL.DETR.ENC_LAYERS
dec_layers = cfg.MODEL.DETR.DEC_LAYERS
pre_norm = cfg.MODEL.DETR.PRE_NORM
# Loss parameters:
giou_weight = cfg.MODEL.DETR.GIOU_WEIGHT
l1_weight = cfg.MODEL.DETR.L1_WEIGHT
deep_supervision = cfg.MODEL.DETR.DEEP_SUPERVISION
no_object_weight = cfg.MODEL.DETR.NO_OBJECT_WEIGHT
N_steps = hidden_dim // 2
d2_backbone = MaskedBackbone(cfg)
backbone = Joiner(d2_backbone,
PositionEmbeddingSine(N_steps, normalize=True))
backbone.num_channels = d2_backbone.num_channels
transformer = Transformer(
d_model=hidden_dim,
dropout=dropout,
nhead=nheads,
dim_feedforward=dim_feedforward,
num_encoder_layers=enc_layers,
num_decoder_layers=dec_layers,
normalize_before=pre_norm,
return_intermediate_dec=deep_supervision,
)
self.detr = DETR(backbone,
transformer,
num_classes=self.num_classes,
num_queries=num_queries,
aux_loss=deep_supervision)
if self.mask_on:
frozen_weights = cfg.MODEL.DETR.FROZEN_WEIGHTS
if frozen_weights != '':
print("LOAD pre-trained weights")
weight = torch.load(
frozen_weights,
map_location=lambda storage, loc: storage)['model']
new_weight = {}
for k, v in weight.items():
if 'detr.' in k:
new_weight[k.replace('detr.', '')] = v
else:
print(f"Skipping loading weight {k} from frozen model")
del weight
self.detr.load_state_dict(new_weight)
del new_weight
self.detr = DETRsegm(self.detr, freeze_detr=(frozen_weights != ''))
self.seg_postprocess = PostProcessSegm
self.detr.to(self.device)
# building criterion
matcher = HungarianMatcher(cost_class=1,
cost_bbox=l1_weight,
cost_giou=giou_weight)
weight_dict = {"loss_ce": 1, "loss_bbox": l1_weight}
weight_dict["loss_giou"] = giou_weight
if deep_supervision:
aux_weight_dict = {}
for i in range(dec_layers - 1):
aux_weight_dict.update(
{k + f"_{i}": v
for k, v in weight_dict.items()})
weight_dict.update(aux_weight_dict)
losses = ["labels", "boxes", "cardinality"]
if self.mask_on:
losses += ["masks"]
self.criterion = SetCriterion(
self.num_classes,
matcher=matcher,
weight_dict=weight_dict,
eos_coef=no_object_weight,
losses=losses,
)
self.criterion.to(self.device)
pixel_mean = torch.Tensor(cfg.MODEL.PIXEL_MEAN).to(self.device).view(
3, 1, 1)
pixel_std = torch.Tensor(cfg.MODEL.PIXEL_STD).to(self.device).view(
3, 1, 1)
self.normalizer = lambda x: (x - pixel_mean) / pixel_std
self.to(self.device)
backbone.num_channels = base_backbone.num_channels
transformer = Transformer(
d_model=args.hidden_dim,
dropout=args.dropout,
nhead=args.nheads,
dim_feedforward=args.dim_feedforward,
num_encoder_layers=args.enc_layers,
num_decoder_layers=args.dec_layers,
normalize_before=args.pre_norm,
return_intermediate_dec=True,
)
model = DETR(
backbone,
transformer,
num_classes=args.num_classes,
num_queries=args.num_queries,
aux_loss=args.aux_loss,
)
matcher = HungarianMatcher(cost_class=args.set_cost_class, cost_bbox=args.set_cost_bbox, cost_giou=args.set_cost_giou)
weight_dict = {'loss_ce': 1, 'loss_bbox': args.bbox_loss_coef}
weight_dict['loss_giou'] = args.giou_loss_coef
if args.aux_loss:
aux_weight_dict = {}
for i in range(args.dec_layers - 1):
aux_weight_dict.update({k + f'_{i}': v for k, v in weight_dict.items()})
weight_dict.update(aux_weight_dict)
losses = ['labels', 'boxes', 'cardinality']
criterion = SetCriterion(args.num_classes, matcher=matcher, weight_dict=weight_dict,
def build_model(args):
num_classes = 20 if args.dataset_file != 'coco' else 91
if args.dataset_file == "coco_panoptic":
num_classes = 250
if args.dataset_file == "ImageNet":
num_classes = 2 # feel free to this num_classes, positive integer larger than 1 is OK.
device = torch.device(args.device)
backbone = build_backbone(args)
transformer = build_transformer(args)
if args.dataset_file == "ImageNet":
model = UPDETR(backbone,
transformer,
num_classes=num_classes,
num_queries=args.num_queries,
aux_loss=args.aux_loss,
num_patches=args.num_patches,
feature_recon=args.feature_recon,
query_shuffle=args.query_shuffle)
else:
model = DETR(
backbone,
transformer,
num_classes=num_classes,
num_queries=args.num_queries,
aux_loss=args.aux_loss,
)
if args.masks:
model = DETRsegm(model,
freeze_detr=(args.frozen_weights is not None))
matcher = build_matcher(args)
weight_dict = {
'loss_ce': 1,
'loss_bbox': args.bbox_loss_coef,
'loss_giou': args.giou_loss_coef
}
if args.dataset_file == 'ImageNet' and args.feature_recon:
weight_dict['loss_feature'] = 1
if args.masks:
weight_dict["loss_mask"] = args.mask_loss_coef
weight_dict["loss_dice"] = args.dice_loss_coef
# TODO this is a hack
if args.aux_loss:
aux_weight_dict = {}
for i in range(args.dec_layers - 1):
aux_weight_dict.update(
{k + f'_{i}': v
for k, v in weight_dict.items()})
weight_dict.update(aux_weight_dict)
losses = ['labels', 'boxes', 'cardinality']
if args.dataset_file == 'ImageNet' and args.feature_recon:
losses += ['feature']
if args.masks:
losses += ["masks"]
criterion = SetCriterion(num_classes,
matcher=matcher,
weight_dict=weight_dict,
eos_coef=args.eos_coef,
losses=losses)
criterion.to(device)
postprocessors = {'bbox': PostProcess()}
if args.masks:
postprocessors['segm'] = PostProcessSegm()
if args.dataset_file == "coco_panoptic":
is_thing_map = {i: i <= 90 for i in range(201)}
postprocessors["panoptic"] = PostProcessPanoptic(is_thing_map,
threshold=0.85)
return model, criterion, postprocessors
