SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.15083 · CVPR · 2023

UniDistill: A Universal Cross-Modality Knowledge Distillation Framework for 3D Object Detection in Bird's-Eye View

Chao Ma, Shuchang Zhou, Chen Hu, Shengchao Zhou, Weizhou Liu

arXiv · PDF · Open in the Atlas

Code that ran

3 of 3 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; 2 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 1 failure of our instrument; 0 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/2303.15083.json · badge: /api/ran/2303.15083.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
megvii-research/cvpr2023-unidistill marked official in Papers with Code
Papers with Code archive
3 ran · 0 no recorded run
FunctionStatusWhere it lives
FeatureDistillLoss Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
megvii-research/cvpr2023-unidistill/unidistill/exps/multisensor_fusion/nuscenes/BEVFusion/BEVFusion_nuscenes_centerhead_camera_exp_distill_fusion.py
code served (permissive licence) · get_code("6146369bb4aae695")
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) and returned Tensor, ndim 0, dtype family float, shape () 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
import torch.nn as nn

torch.manual_seed(0)

# Analyze the function to understand tensor shapes:
# - feature_lidar and feature_fuse are used with grid_sample, so they need shape (batch, channels, height, width)
# - gt_boxes_bev_coords is indexed and has mean taken over dim=2, then unsqueeze(2), so shape (batch, num_boxes, 4, 2)
# - gt_boxes_indices is used for boolean indexing on sampled features, shape (batch, num_boxes)

batch_size = 2
num_boxes = 3
channels = 4
height = 8
width = 6

# feature_lidar: (batch, channels, height, width)
feature_lidar = torch.randn(batch_size, channels, height, width)

# feature_fuse: (batch, channels, height, width)
feature_fuse = torch.randn(batch_size, channels, height, width)

# gt_boxes_bev_coords: (batch, num_boxes, 4, 2)
# 4 corners of bounding box, 2 coordinates (x, y)
gt_boxes_bev_coords = torch.randn(batch_size, num_boxes, 4, 2) * 2 - 1  # normalized to roughly [-1, 1]

# gt_boxes_indices: (batch, num_boxes) boolean tensor
# Used to index into sampled features
gt_boxes_indices = torch.randint(0, 2, (batch_size, num_boxes)).bool()

ARGS = {
    "feature_lidar": feature_lidar,
    "feature_fuse": feature_fuse,
    "gt_boxes_bev_coords": gt_boxes_bev_coords,
    "gt_boxes_indices": gt_boxes_indices,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 0, dtype family float, shape () — shape () 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
reduce_mean Ran
ran, honoured the contract we drafted
megvii-research/cvpr2023-unidistill/unidistill/exps/multisensor_fusion/nuscenes/BEVFusion/BEVFusion_nuscenes_centerhead_camera_exp_distill_fusion.py
code served (permissive licence) · get_code("11790e16efd1a8d4")
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 torch
import torch.nn as nn
from torch import distributed as dist

# Since distributed is not initialized in this context,
# the functions will return early and just work with the tensor directly.
# We need to provide a tensor that the functions can operate on.

torch.manual_seed(0)

# Create a small tensor for reduce_mean to operate on
tensor = torch.randn(2, 4)

ARGS = {
    "tensor": tensor
}
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, 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
reduce_sum Ran
ran, honoured the contract we drafted
megvii-research/cvpr2023-unidistill/unidistill/exps/multisensor_fusion/nuscenes/BEVFusion/BEVFusion_nuscenes_centerhead_camera_exp_distill_fusion.py
code served (permissive licence) · get_code("350276fe38eb1d9f")
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 torch

torch.manual_seed(0)

# The function reduce_sum takes a tensor and optionally uses distributed operations
# Since dist.is_initialized() will return False in a non-distributed setting,
# the function will simply return the tensor (or a clone of it)
# We need to provide a valid tensor argument

tensor = torch.randn(2, 4)

ARGS = {
    "tensor": tensor,
}
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, 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

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 3 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.

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

In the field of 3D object detection for autonomous driving, the sensor portfolio including multi-modality and single-modality is diverse and complex. Since the multimodal methods have system complexity while the accuracy of single-modal ones is relatively low, how to make a tradeoff between them is difficult. In this work, we propose a universal cross-modality knowledge distillation framework (UniDistill) to improve the performance of single-modality detectors. Specifically, during training, UniDistill projects the features of both the teacher and the student detector into Bird's-Eye-View (BEV), which is a friendly representation for different modalities. Then, three distillation losses are calculated to sparsely align the foreground features, helping the student learn from the teacher without introducing additional cost during inference. Taking advantage of the similar detection paradigm of different detectors in BEV, UniDistill easily supports LiDAR-to-camera, camera-to-LiDAR, fusion-to-LiDAR and fusion-to-camera distillation paths. Furthermore, the three distillation losses can filter the effect of misaligned background information and balance between objects of different sizes, improving the distillation effectiveness. Extensive experiments on nuScenes demonstrate that UniDistill effectively improves the mAP and NDS of student detectors by 2.0%∼3.2%.

For agents

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

get_harvested_code_for_paper("2303.15083")
get_code_for_paper("2303.15083")
have("2303.15083")

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

curl https://syntology.ai/api/ran/2303.15083.json

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

[![Syntology run record](https://syntology.ai/api/ran/2303.15083.svg)](https://syntology.ai/paper/2303.15083)

Connect an agent — have() is free.