-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathNOTICE
More file actions
46 lines (38 loc) · 2.21 KB
/
NOTICE
File metadata and controls
46 lines (38 loc) · 2.21 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
CIR-Phys
Copyright 2024 Max-Planck-Gesellschaft
Licensed under the MIT License (see LICENSE)
Code author: Michael Strecke, michael.strecke@tuebingen.mpg.de
Embodied Vision Group, Max Planck Institute for Intelligent Systems
-------------------
The code in this repository is based on the CIR repository at
https://github.com/princeton-vl/Coupled-Iterative-Refinement/tree/c50df7816714007c7f2f5188995807b3b396ad3d,
licensed under the MIT License (see CIR-LICENSE).
Additions/modifications to existing code were made in the following locations:
- demo_batch.py and test_batch.py are based on demo.py and test.py from the repository mentioned above. Changes were
made to batch-process all detections in an image at once instead of iterating over the object detections. Additionally,
flags for activating the intersection constraint, using ground truth masks, and processing the synpick dataset were
added.
- The following source files were added compared to the CIR repository:
convert_mesh_to_sdf.py
additional_scripts/run_gcpr24_eval.py
additional_scripts/eval_penetration_errors.py
additional_scripts/eval_penetration_scores.py
additional_scripts/splits_to_targets.py
cir/utils/geom/intersection_ops.py
- All other source files originate from the CIR repository. We modified the following files (see the files for details
on the modifications):
cir/datasets/bop.py
cir/pose_models/raftse3.py
cir/utils/geom/ba.py
cir/utils/augmentation.py
cir/utils/pt_renderer.py
The following files in the CIR repository were adapted from https://github.com/lukemelas/EfficientNet-PyTorch by the CIR
authors. The original code is licensed under the Apache License, Version 2.0 (see EfficientNet-LICENSE).
cir/pose_models/efficientnet.py
cir/pose_models/efficientnet_utils.py
-------------------
For evaluation, we rely on code from the bop toolkit at
https://github.com/thodan/bop_toolkit/tree/37d79c4c5fb027da92bc40f36b82ea9b7b197f1d, licensed under the MIT license
(see https://github.com/thodan/bop_toolkit/tree/37d79c4c5fb027da92bc40f36b82ea9b7b197f1d/LICENSE)
The file additional_scripts/bop_toolkit.patch is intended to patch the bop toolkit repo at the specified commit and
contains code snippets from the repository.