SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2104.02323 · CVPR · 2021

Objects are Different: Flexible Monocular 3D Object Detection

Jiwen Lu, Jie Zhou, Yunpeng Zhang

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 20 functions out of this paper's own repositories and ran 11 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
zhangyp15/MonoFlex canonical 2 of 5
Owen-Liuyuxuan/visualDet3D — 9 of 15
FunctionStatusWhere it lives
ClipBoxes Ran Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("570e350e7fe7526b")
IoULoss Ran Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("c5197ebb950e466e")
MonoFlexHead Ran Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("8e7a3c3289cb3e80")
_nms Ran Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("59022d9630576578")
boxes3d_to_bev_torch Ran Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("af38ff2bb862b433")
cat Ran zhangyp15/MonoFlex/model/utils.py
code served (permissive licence) · get_code("6cbf46794989b731")
compute_bin_loss Ran Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("65d7a40d70c53ed9")
compute_res_loss Ran Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("7df13fb9e54d592b")
decode_depth_from_keypoints Ran Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("74be68131587e23b")
decode_depth_inv_sigmoid Ran Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("4bf22db9eaf70901")
strip_prefix_if_present Ran zhangyp15/MonoFlex/utils/model_serialization.py
code served (permissive licence) · get_code("5a951495ac113177")
KM3DHead Not yet run Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("7b2125261c5b536d")
Position_loss Not yet run Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("8469c42e1fc712db")
_topk_channel Not yet run Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("03b36f268d8645d0")
affine_transform Not yet run zhangyp15/MonoFlex/model/heatmap_coder.py
code served (permissive licence) · get_code("65a16c264996059d")
boxes_iou3d_gpu Not yet run Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("a7fa48606189776e")
compute_rot_loss Not yet run Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("9d749234d255406c")
gaussian_radius Not yet run zhangyp15/MonoFlex/model/heatmap_coder.py
code served (permissive licence) · get_code("c044d44b5a7087e4")
gen_position Not yet run Owen-Liuyuxuan/visualDet3D/visualDet3D/networks/heads/monoflex_head.py
code served (permissive licence) · get_code("dae0c091d06a2048")
get_transfrom_matrix Not yet run zhangyp15/MonoFlex/model/heatmap_coder.py
code served (permissive licence) · get_code("194052ec40659a42")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

The precise localization of 3D objects from a single image without depth information is a highly challenging problem. Most existing methods adopt the same approach for all objects regardless of their diverse distributions, leading to limited performance for truncated objects. In this paper, we propose a flexible framework for monocular 3D object detection which explicitly decouples the truncated objects and adaptively combines multiple approaches for object depth estimation. Specifically, we decouple the edge of the feature map for predicting long-tail truncated objects so that the optimization of normal objects is not influenced. Furthermore, we formulate the object depth estimation as an uncertainty-guided ensemble of directly regressed object depth and solved depths from different groups of keypoints. Experiments demonstrate that our method outperforms the state-of-the-art method by relatively 27% for the moderate level and 30% for the hard level in the test set of KITTI benchmark while maintaining real-time efficiency. Code will be available at https://github. com/zhangyp15/MonoFlex.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2104.02323")
get_code_for_paper("2104.02323")
have("2104.02323")

Connect an agent — have() is free.