Skip to content

Fix dependencies; Accept focal as input; Return detection confidence #21

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
alex-bene wants to merge 4 commits into
base: main
Choose a base branch
from
Open

Fix dependencies; Accept focal as input; Return detection confidence #21

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
c2d1ef6
relax torch dep; update ultralytics dep
alex-bene May 15, 2026
9766463
focal_length as inference input
alex-bene May 15, 2026
99d41de
expose detection confidence from YOLO to the return dict
alex-bene May 18, 2026
cf8f582
use chumpy pr with install fix
alex-bene May 22, 2026
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
6 changes: 3 additions & 3 deletions requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,11 +1,11 @@
torch>=2.0,<=2.5
torch>=2.0
torchvision>=0.15.1
smplx==0.1.28
timm
einops
ultralytics==8.1.34
ultralytics>=8.3.4
opencv-python
huggingface_hub
scikit-image
roma
chumpy @ git+https://github.com/mattloper/chumpy
chumpy @ git+https://github.com/mattloper/chumpy@4228d703b622e172e843438fe0fada102979361a
10 changes: 7 additions & 3 deletions wilor_mini/pipelines/wilor_hand_pose3d_estimation_pipeline.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -86,7 +86,11 @@ def predict(self, image, **kwargs):
hand_bbox = det.boxes.data.cpu().detach().squeeze().numpy()
is_rights.append(det.boxes.cls.cpu().detach().squeeze().item())
bboxes.append(hand_bbox[:4].tolist())
detect_rets.append({"hand_bbox": bboxes[-1], "is_right": is_rights[-1]})
detect_rets.append({
"hand_bbox": bboxes[-1],
"is_right": is_rights[-1],
"hand_conf": det.boxes.conf.cpu().detach().squeeze().item(),
})

if len(bboxes) == 0:
self.logger.warn("No hand detected!")
Expand Down Expand Up @@ -144,7 +148,7 @@ def predict(self, image, **kwargs):
wilor_output_i["hand_pose"] = np.concatenate(
(wilor_output_i["hand_pose"][:, :, 0:1], -wilor_output_i["hand_pose"][:, :, 1:3]),
axis=-1)
scaled_focal_length = self.FOCAL_LENGTH / self.IMAGE_SIZE * img_size.max()
scaled_focal_length = kwargs.get("focal_length", self.FOCAL_LENGTH / self.IMAGE_SIZE * img_size.max())
pred_cam_t_full = utils.cam_crop_to_full(pred_cam, box_center[None], bbox_size, img_size[None],
scaled_focal_length)
wilor_output_i["pred_cam_t_full"] = pred_cam_t_full
Expand Down Expand Up @@ -221,7 +225,7 @@ def predict_with_bboxes(self, image, bboxes, is_rights, **kwargs):
wilor_output_i["hand_pose"] = np.concatenate(
(wilor_output_i["hand_pose"][:, :, 0:1], -wilor_output_i["hand_pose"][:, :, 1:3]),
axis=-1)
scaled_focal_length = self.FOCAL_LENGTH / self.IMAGE_SIZE * img_size.max()
scaled_focal_length = kwargs.get("focal_length", self.FOCAL_LENGTH / self.IMAGE_SIZE * img_size.max())
pred_cam_t_full = utils.cam_crop_to_full(pred_cam, box_center[None], bbox_size, img_size[None],
scaled_focal_length)
wilor_output_i["pred_cam_t_full"] = pred_cam_t_full
Expand Down