# setup backend
be = gen_backend(**extract_valid_args(args, gen_backend))
be.enable_winograd = 4 # default to winograd 4 for fast autotune
# directory to store VGG weights
cache_dir = get_data_cache_dir(args.data_dir, subdir='pascalvoc_cache')
# build data loader
# get config file for PASCALVOC
config = PASCALVOC(args.manifest['train'], args.manifest_root,
width=args.width, height=args.height,
rois_per_img=rpn_rois_per_img, inference=False)
config['subset_fraction'] = float(args.subset_pct / 100.0)
train_set = faster_rcnn.build_dataloader(config, frcn_rois_per_img)
# build the Faster-RCNN model
model = faster_rcnn.build_model(train_set, frcn_rois_per_img, inference=False)
# set up cost different branches, respectively
weights = 1.0 / (rpn_rois_per_img)
roi_w = 1.0 / (frcn_rois_per_img)
frcn_tree_cost = Multicost(costs=[GeneralizedCostMask(costfunc=CrossEntropyMulti(), weights=roi_w),
GeneralizedCostMask(costfunc=SmoothL1Loss(), weights=roi_w)
], weights=[1, 1])
cost = Multicost(costs=[GeneralizedCostMask(costfunc=CrossEntropyMulti(), weights=weights),
GeneralizedCostMask(costfunc=SmoothL1Loss(sigma=3.0), weights=weights),
frcn_tree_cost,
# hyperparameters
assert args.batch_size is 1, "Faster-RCNN only supports batch size 1"
rpn_rois_per_img = 256
frcn_rois_per_img = 128
# setup backend
be = gen_backend(**extract_valid_args(args, gen_backend))
# build data loader
cache_dir = get_data_cache_dir(args.data_dir, subdir='pascalvoc_cache')
config = PASCALVOC(args.manifest['val'], args.manifest_root,
width=args.width, height=args.height,
rois_per_img=rpn_rois_per_img, inference=True)
valid_set = faster_rcnn.build_dataloader(config, be, frcn_rois_per_img)
num_classes = valid_set.num_classes
# build the Faster-RCNN network
(model, proposalLayer) = faster_rcnn.build_model(valid_set, frcn_rois_per_img, inference=True)
# load parameters and initialize model
model.load_params(args.model_file)
model.initialize(dataset=valid_set)
# normalize the model by the bbtarget mean and std if needed
# if a full training run was completed using train.py, then normalization
# was already performed prior to saving the model.
if args.normalize:
model = util.scale_bbreg_weights(model, [0.0, 0.0, 0.0, 0.0],
rpn_rois_per_img = 256
frcn_rois_per_img = 128
# setup backend
be = gen_backend(**extract_valid_args(args, gen_backend))
# build data loader
cache_dir = get_data_cache_dir(args.data_dir, subdir='pascalvoc_cache')
config = PASCALVOC(args.manifest['val'],
args.manifest_root,
width=args.width,
height=args.height,
rois_per_img=rpn_rois_per_img,
inference=True)
valid_set = faster_rcnn.build_dataloader(config, frcn_rois_per_img)
num_classes = valid_set.num_classes
# build the Faster-RCNN network
(model, proposalLayer) = faster_rcnn.build_model(valid_set,
frcn_rois_per_img,
inference=True)
# load parameters and initialize model
model.load_params(args.model_file, load_states=False)
model.initialize(dataset=valid_set)
# normalize the model by the bbtarget mean and std if needed
# if a full training run was completed using train.py, then normalization
# was already performed prior to saving the model.
be.enable_winograd = 4 # default to winograd 4 for fast autotune
# directory to store VGG weights
cache_dir = get_data_cache_dir(args.data_dir, subdir='pascalvoc_cache')
# build data loader
# get config file for PASCALVOC
config = PASCALVOC(args.manifest['train'],
args.manifest_root,
width=args.width,
height=args.height,
rois_per_img=rpn_rois_per_img,
inference=False)
config['subset_fraction'] = float(args.subset_pct / 100.0)
train_set = faster_rcnn.build_dataloader(config, be, frcn_rois_per_img)
# build the Faster-RCNN model
model = faster_rcnn.build_model(train_set, frcn_rois_per_img, inference=False)
# set up cost different branches, respectively
weights = 1.0 / (rpn_rois_per_img)
roi_w = 1.0 / (frcn_rois_per_img)
frcn_tree_cost = Multicost(costs=[
GeneralizedCostMask(costfunc=CrossEntropyMulti(), weights=roi_w),
GeneralizedCostMask(costfunc=SmoothL1Loss(), weights=roi_w)
],
weights=[1, 1])
cost = Multicost(costs=[