Skip to content

fix cliff inference problems #342

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into
base: main
Choose a base branch
from
Open

fix cliff inference problems #342

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
1d35c71
fix cliff inference problems
Apr 18, 2023
50824c7
fix yapf check
Apr 18, 2023
b56c62e
fix config yapf check
Apr 18, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 14 additions & 5 deletions configs/cliff/resnet50_pw3d_cache.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,7 @@
# dict(type='TensorboardLoggerHook')
])

img_resolution = (192, 256)
img_res = (192, 256)

# model settings
model = dict(
Expand Down Expand Up @@ -94,7 +94,7 @@
dict(type='RandomHorizontalFlip', flip_prob=0.5, convention='smpl_54'),
dict(type='GetRandomScaleRotation', rot_factor=30, scale_factor=0.25),
dict(type='GetBboxInfo'),
dict(type='MeshAffine', img_res=img_resolution),
dict(type='MeshAffine', img_res=img_res),
dict(type='Normalize', **img_norm_cfg),
dict(type='ImageToTensor', keys=['img']),
dict(type='ToTensor', keys=data_keys),
Expand All @@ -111,7 +111,7 @@
dict(type='LoadImageFromFile'),
dict(type='GetRandomScaleRotation', rot_factor=0, scale_factor=0),
dict(type='GetBboxInfo'),
dict(type='MeshAffine', img_res=img_resolution),
dict(type='MeshAffine', img_res=img_res),
dict(type='Normalize', **img_norm_cfg),
dict(type='ImageToTensor', keys=['img']),
dict(type='ToTensor', keys=data_keys),
Expand All @@ -125,12 +125,21 @@
]

inference_pipeline = [
dict(type='MeshAffine', img_res=img_resolution),
dict(type='GetBboxInfo'),
dict(type='MeshAffine', img_res=img_res),
dict(type='Normalize', **img_norm_cfg),
dict(type='ImageToTensor', keys=['img']),
dict(
type='ToTensor',
keys=[
'bbox_info', 'img_h', 'img_w', 'center', 'scale', 'focal_length'
]),
dict(
type='Collect',
keys=['img', 'sample_idx'],
keys=[
'img', 'sample_idx', 'bbox_info', 'img_h', 'img_w', 'center',
'scale', 'focal_length'
],
meta_keys=['image_path', 'center', 'scale', 'rotation'])
]

Expand Down
16 changes: 7 additions & 9 deletions mmhuman3d/apis/inference.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -183,22 +183,20 @@ def inference_image_based_model(

batch_data = collate(batch_data, samples_per_gpu=1)

if next(model.parameters()).is_cuda:
# scatter not work so just move image to cuda device
batch_data['img'] = batch_data['img'].to(device)

# get all img_metas of each bounding box
batch_data['img_metas'] = [
img_metas[0] for img_metas in batch_data['img_metas'].data
]
if next(model.parameters()).is_cuda:
# scatter not work so just move image to cuda device
batch_data = dict(
map(
lambda item: item if not isinstance(item[1], torch.Tensor) else
(item[0], item[1].to(device)), batch_data.items()))

# forward the model
with torch.no_grad():
results = model(
img=batch_data['img'],
img_metas=batch_data['img_metas'],
sample_idx=batch_data['sample_idx'],
)
results = model(**batch_data)

for idx in range(len(det_results)):
mesh_result = det_results[idx].copy()
Expand Down