SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.12396 · CVPR · 2023

Rigidity-Aware Detection for 6D Object Pose Estimation

Rui Song, Mathieu Salzmann, Yinlin Hu, Clearspace, Yang Hai, Jiaojiao Li

arXiv · PDF · Open in the Atlas

Code that ran

6 of 9 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 4 failures of our instrument; 3 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2303.12396.json · badge: /api/ran/2303.12396.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
YangHai-1218/RADet canonical (classifier label)
classifier basis not recorded on the link
6 ran · 3 no recorded run
FunctionStatusWhere it lives
anchor_inside_flags Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
YangHai-1218/RADet/radet/core/anchor/utils.py
code served (permissive licence) · get_code("cc1d87b0a0678212")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 1, dtype family float, shape (2,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create flat_anchors: shape (n, 4) representing anchor boxes [x1, y1, x2, y2]
flat_anchors = torch.tensor([
    [10.0, 20.0, 50.0, 80.0],
    [5.0, 15.0, 100.0, 120.0]
], dtype=torch.float32)

# Create valid_flags: boolean tensor of same length as flat_anchors
valid_flags = torch.tensor([True, True], dtype=torch.bool)

# Image shape (height, width, channels) - only first 2 elements used
img_shape = (256, 512, 3)

# allowed_border parameter
allowed_border = 0

ARGS = {
    'flat_anchors': flat_anchors,
    'valid_flags': valid_flags,
    'img_shape': img_shape,
    'allowed_border': allowed_border
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 1, dtype family float, shape (2,) — shape (2,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
bbox_flip Ran
ran on a synthesized input
YangHai-1218/RADet/radet/core/bbox/transforms.py
code served (permissive licence) · get_code("6a29f5a7da19f7ae")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 8) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create bboxes tensor with shape (..., 4*k)
# Using shape (2, 8) means batch of 2, with 2 bboxes (4*2=8 coordinates)
bboxes = torch.tensor([
    [10.0, 20.0, 30.0, 40.0, 50.0, 60.0, 70.0, 80.0],
    [15.0, 25.0, 35.0, 45.0, 55.0, 65.0, 75.0, 85.0]
])

# Image shape as (height, width)
img_shape = (100, 120)

# Direction for flipping
direction = 'horizontal'

ARGS = {
    'bboxes': bboxes,
    'img_shape': img_shape,
    'direction': direction
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 8) — shape (2, 8) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
bbox_overlaps Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
YangHai-1218/RADet/radet/core/evaluation/bbox_overlaps.py
code served (permissive licence) · get_code("c21eb325c33ec5d7")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 2, dtype family float, shape (2, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create small bounding boxes for testing
# Format: [x1, y1, x2, y2]
bboxes1 = np.array([
    [10, 10, 50, 50],
    [20, 20, 60, 60]
], dtype=np.int64)

bboxes2 = np.array([
    [30, 30, 70, 70],
    [40, 40, 80, 80],
    [5, 5, 25, 25]
], dtype=np.int64)

ARGS = {
    'bboxes1': bboxes1,
    'bboxes2': bboxes2,
    'mode': 'iou',
    'eps': 1e-06
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (2, 3) — shape (2, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
ensure_rng Ran
ran; our contract draft was wrong, not the code
YangHai-1218/RADet/radet/core/bbox/demodata.py
code served (permissive licence) · get_code("272ba15451a11158")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned RandomState. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Test case 1: None argument (uses global rng)
args_none = None

# Test case 2: Integer seed
args_int = 42

# Test case 3: RandomState object
args_randomstate = np.random.RandomState(123)

# Create ARGS dict with all parameter combinations
ARGS = {
    'rng': None
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type RandomState — a RandomState output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a RandomState output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
images_to_levels Ran
ran; our contract draft was wrong, not the code
YangHai-1218/RADet/radet/core/anchor/utils.py
code served (permissive licence) · get_code("0910773f52297bf8")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 3, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Analyze the function:
# - target: a list/sequence that gets stacked into a tensor with torch.stack
# - num_levels: a sequence of integers representing the number of levels and their sizes
# 
# The function stacks targets by image, then splits them by levels.
# target should be a list of tensors (one per image)
# num_levels should be a sequence of integers that sum to the feature dimension

# Create a list of target tensors (one per image)
# Each tensor represents features for that image
# Let's say we have 2 images, and features of size 4
target = [
    torch.randn(4),  # image 0: 4 features
    torch.randn(4),  # image 1: 4 features
]

# num_levels specifies how to split the 4 features into levels
# For example: [2, 2] means level 0 gets first 2 features, level 1 gets next 2
num_levels = [2, 2]

ARGS = {
    "target": target,
    "num_levels": num_levels,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 3, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
random_boxes Ran
ran, honoured the contract we drafted
YangHai-1218/RADet/radet/core/bbox/demodata.py
code served (permissive licence) · get_code("00430ee7a62da222")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import torch

torch.manual_seed(0)

ARGS = {
    "num": 2,
    "scale": 512,
    "rng": 42,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
bbox_mapping Not yet run
unverified (harvested, no recorded run)
YangHai-1218/RADet/radet/core/bbox/transforms.py
code served (permissive licence) · get_code("151f19f8cd22a341")
How it ran

no invocation recorded.

bbox_mapping_back Not yet run
unverified (harvested, no recorded run)
YangHai-1218/RADet/radet/core/bbox/transforms.py
code served (permissive licence) · get_code("994aebbe25cce42d")
How it ran

no invocation recorded.

calc_region Not yet run
unverified (harvested, no recorded run)
YangHai-1218/RADet/radet/core/anchor/utils.py
code served (permissive licence) · get_code("4befdb30b7583175")
How it ran

no invocation recorded.

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 distinct functions with an identity, 9 are also present in other papers whose code we hold (9 with at least one paper dated earlier, 3 the same month, 4 later) and 0 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
anchor_inside_flags 3 other papers
1 dated earlier, 1 the same month, 1 later
1906.07155 (2019-06)
MMDetection: Open MMLab Detection Toolbox and Benchmark
2303.06999 (2023-03)
Identifying Label Errors in Object Detection Datasets by Loss Inspection
2401.09923 (2024-01)
MAMBA: Multi-level Aggregation via Memory Bank for Video Object Detection
calc_region 3 other papers
1 dated earlier, 1 the same month, 1 later
1906.07155 (2019-06)
MMDetection: Open MMLab Detection Toolbox and Benchmark
2303.06999 (2023-03)
Identifying Label Errors in Object Detection Datasets by Loss Inspection
2401.09923 (2024-01)
MAMBA: Multi-level Aggregation via Memory Bank for Video Object Detection
images_to_levels 3 other papers
1 dated earlier, 1 the same month, 1 later
1906.07155 (2019-06)
MMDetection: Open MMLab Detection Toolbox and Benchmark
2303.06999 (2023-03)
Identifying Label Errors in Object Detection Datasets by Loss Inspection
2401.09923 (2024-01)
MAMBA: Multi-level Aggregation via Memory Bank for Video Object Detection
bbox_overlaps 2 other papers
1 dated earlier, 1 later
1906.07155 (2019-06)
MMDetection: Open MMLab Detection Toolbox and Benchmark
2407.07402 (2024-07)
ActionVOS: Actions as Prompts for Video Object Segmentation
ensure_rng 2 other papers
2 dated earlier
1906.07155 (2019-06)
MMDetection: Open MMLab Detection Toolbox and Benchmark
2004.06002 (2020-04)
Dynamic R-CNN: Towards High Quality Object Detection via Dynamic Training
random_boxes 2 other papers
2 dated earlier
1906.07155 (2019-06)
MMDetection: Open MMLab Detection Toolbox and Benchmark
2004.06002 (2020-04)
Dynamic R-CNN: Towards High Quality Object Detection via Dynamic Training
bbox_flip 1 other paper
1 dated earlier
1906.07155 (2019-06)
MMDetection: Open MMLab Detection Toolbox and Benchmark
bbox_mapping 1 other paper
1 dated earlier
1906.07155 (2019-06)
MMDetection: Open MMLab Detection Toolbox and Benchmark
bbox_mapping_back 1 other paper
1 dated earlier
1906.07155 (2019-06)
MMDetection: Open MMLab Detection Toolbox and Benchmark

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

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

Abstract

Most recent 6D object pose estimation methods first use object detection to obtain 2D bounding boxes before actually regressing the pose. However, the general object detection methods they use are ill-suited to handle cluttered scenes, thus producing poor initialization to the subsequent pose network. To address this, we propose a rigidity-aware detection method exploiting the fact that, in 6D pose estimation, the target objects are rigid. This lets us introduce an approach to sampling positive object regions from the entire visible object area during training, instead of naively drawing samples from the bounding box center where the object might be occluded. As such, every visible object part can contribute to the final bounding box prediction, yielding better detection robustness. Key to the success of our approach is a visibility map, which we propose to build using a minimum barrier distance between every pixel in the bounding box and the box boundary. Our results on seven challenging 6D pose estimation datasets evidence that our method outperforms general detection frameworks by a large margin. Furthermore, combined with a pose regression network, we obtain state-of-the-art pose estimation results on the challenging BOP benchmark.

For agents

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

get_harvested_code_for_paper("2303.12396")
get_code_for_paper("2303.12396")
have("2303.12396")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2303.12396.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2303.12396.svg)](https://syntology.ai/paper/2303.12396)

Connect an agent — have() is free.