SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2409.12753 · 2024

DrivingForward: Feed-forward 3D Gaussian Splatting for Driving Scene Reconstruction from Flexible Surround-view Input

arXiv · PDF · Open in the Atlas

Code that ran

9 of 11 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; 3 honoured the contract we drafted, 3 violated it and 0 have no contract verdict; 3 failures of our instrument; 2 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/2409.12753.json · badge: /api/ran/2409.12753.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
fangzhou2000/drivingforward canonical (classifier label)
classifier basis: census
9 ran · 2 no recorded run
FunctionStatusWhere it lives
get_adj_cams Ran
ran, honoured the contract we drafted
fangzhou2000/drivingforward/models/gaussian/GaussianRender.py
code served (permissive licence) · get_code("dd34eb67d4d8d0c6")
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 1, dtype family int; 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)

left_cam_dict = {2: 0, 0: 1, 4: 2, 1: 3, 5: 4, 3: 5}
right_cam_dict = {0: 2, 1: 0, 2: 4, 3: 1, 4: 5, 5: 3}

def get_adj_cams(cam):
    adj_cams = [cam]
    adj_cams.append(left_cam_dict[cam])
    adj_cams.append(right_cam_dict[cam])
    return adj_cams

ARGS = {
    "cam": 0
}
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 1, dtype family int — 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
img_loader Ran
ran, honoured the contract we drafted
fangzhou2000/drivingforward/dataset/data_util.py
code served (permissive licence) · get_code("8c6e232d296289b0")
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 Image. 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 PIL.Image as pil
import io
import tempfile
import os

# Create a temporary RGB image file
img_array = pil.new('RGB', (10, 10), color=(255, 0, 0))
temp_dir = tempfile.mkdtemp()
temp_path = os.path.join(temp_dir, "test_image.png")
img_array.save(temp_path)

ARGS = {
    'path': temp_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Image — a Image output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Image output the recorded ndim 3 and dtype family int 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
is_list Ran
ran, violated the contract we drafted
fangzhou2000/drivingforward/dataset/nuscenes_dataset.py
code served (permissive licence) · get_code("f7264905458095af")
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 bool, 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 numpy as np

torch.manual_seed(0)

# The function is_list simply checks if data is a list
# We need to provide a value for the 'data' parameter

ARGS = {
    "data": [1, 2, 3, 4, 5]
}
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 bool, 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
is_numpy Ran
ran, violated the contract we drafted
fangzhou2000/drivingforward/dataset/nuscenes_dataset.py
code served (permissive licence) · get_code("4c3b87f6a4a1e66f")
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 bool, 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 numpy as np

# Create test data for is_numpy function
numpy_array = np.array([1, 2, 3, 4, 5])

# The function takes one parameter: data
ARGS = {
    "data": numpy_array
}
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 bool, 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
is_tensor Ran
ran, violated the contract we drafted
fangzhou2000/drivingforward/dataset/nuscenes_dataset.py
code served (permissive licence) · get_code("8a89932040fdda13")
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 bool, 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

torch.manual_seed(0)

# The function checks if data is a torch tensor by type comparison
# We need to provide a 'data' parameter

# Create a valid torch tensor
tensor_data = torch.randn(2, 4)

# Create a non-tensor to test the function works both ways
non_tensor_data = [1, 2, 3, 4]

ARGS = {
    "data": tensor_data
}
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 bool, 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
mask_loader_scene Ran
ran, honoured the contract we drafted
fangzhou2000/drivingforward/dataset/data_util.py
code served (permissive licence) · get_code("c817bf0f393ecda3")
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 Image. 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 os
import tempfile
import PIL.Image as pil
import numpy as np

# Create a temporary directory with the expected structure
temp_dir = tempfile.mkdtemp()
mask_idx = 0
cam = "front"

# Create subdirectory for mask_idx
mask_dir = os.path.join(temp_dir, str(mask_idx))
os.makedirs(mask_dir, exist_ok=True)

# Create a simple grayscale PNG mask image
mask_array = np.random.randint(0, 256, (32, 32), dtype=np.uint8)
mask_img = pil.fromarray(mask_array, mode='L')
mask_path = os.path.join(mask_dir, '{}_mask.png'.format(cam.upper()))
mask_img.save(mask_path)

# Create a simple camera object that has an upper() method
class SimpleCamera:
    def __init__(self, name):
        self.name = name
    
    def upper(self):
        return self.name.upper()

camera = SimpleCamera(cam)

ARGS = {
    'path': temp_dir,
    'mask_idx': mask_idx,
    'cam': camera
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Image — a Image output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Image output the recorded ndim 2 and dtype family int 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
pack_cam_feat Ran
ran; our contract draft was wrong, not the code
fangzhou2000/drivingforward/network/blocks.py
code served (permissive licence) · get_code("acecdd618ca22cec")
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 dict. 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 torch
import torch.nn as nn

torch.manual_seed(0)

# Test case 1: tensor input (not dict)
tensor_input = torch.randn(2, 3, 4, 5)  # batch=2, n_cam=3, then other dims

# Test case 2: dict input with multiple keys
dict_input = {
    'features': torch.randn(2, 3, 4, 5),  # batch=2, n_cam=3, height=4, width=5
    'depth': torch.randn(2, 3, 6, 7),     # batch=2, n_cam=3, height=6, width=7
}

# The function accepts either a tensor or a dict of tensors
# We'll test with the dict case as it exercises more code
ARGS = {
    'x': dict_input
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
unpack_cam_feat Ran
ran; our contract draft was wrong, not the code
fangzhou2000/drivingforward/network/blocks.py
code served (permissive licence) · get_code("88e75922582c3bfe")
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 dict. 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 torch

torch.manual_seed(0)

# Case 1: x is a dict of tensors
# The function reshapes each tensor in the dict to (b, n_cam, *remaining_dims)
x_dict = {
    'feat1': torch.randn(4, 12),  # Will be reshaped to (2, 2, 12)
    'feat2': torch.randn(4, 8, 3),  # Will be reshaped to (2, 2, 8, 3)
}

b = 2
n_cam = 2

ARGS = {
    'x': x_dict,
    'b': b,
    'n_cam': n_cam,
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
upsample Ran
ran; our contract draft was wrong, not the code
fangzhou2000/drivingforward/network/blocks.py
code served (permissive licence) · get_code("ba567a41af774393")
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 4, dtype family float, shape (2, 4, 6, 10) 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.functional as F

torch.manual_seed(0)

# x is a tensor that will be interpolated
# F.interpolate expects a 4D tensor (batch, channels, height, width) for 2D upsampling
x = torch.randn(2, 4, 3, 5)

ARGS = {
    'x': x
}
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 4, dtype family float, shape (2, 4, 6, 10) — shape (2, 4, 6, 10) 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
pts2render Not yet run
unverified (harvested, no recorded run)
fangzhou2000/drivingforward/models/gaussian/GaussianRender.py
code served (permissive licence) · get_code("04ee2d7fa7fb6811")
How it ran

no invocation recorded.

transform_mask_sample Not yet run
unverified (harvested, no recorded run)
fangzhou2000/drivingforward/dataset/data_util.py
code served (permissive licence) · get_code("a99c4a3ee761b6c5")
How it ran

no invocation recorded.

Also present in other papers' code

Of 11 samples lifted from this paper's linked repositories, 11 carry a function identity and 0 do not. Of the 11 distinct functions with an identity, 1 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 0 later) and 10 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
upsample 8 other papers
8 dated earlier
1912.03426 (2019-12)
Self-Supervised 3D Keypoint Learning for Ego-motion Estimation
2002.08394 (2020-02)
MonoLayout: Amodal scene layout from a single image
2103.14201 (2021-03)
Image2Reverb: Cross-Modal Reverb Impulse Response Synthesis
and 5 more not named here

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

We propose DrivingForward, a feed-forward Gaussian Splatting model that reconstructs driving scenes from flexible surround-view input. Driving scene images from vehicle-mounted cameras are typically sparse, with limited overlap, and the movement of the vehicle further complicates the acquisition of camera extrinsics. To tackle these challenges and achieve real-time reconstruction, we jointly train a pose network, a depth network, and a Gaussian network to predict the Gaussian primitives that represent the driving scenes. The pose network and depth network determine the position of the Gaussian primitives in a self-supervised manner, without using depth ground truth and camera extrinsics during training. The Gaussian network independently predicts primitive parameters from each input image, including covariance, opacity, and spherical harmonics coefficients. At the inference stage, our model can achieve feed-forward reconstruction from flexible multi-frame surround-view input. Experiments on the nuScenes dataset show that our model outperforms existing state-of-the-art feed-forward and scene-optimized reconstruction methods in terms of reconstruction.

For agents

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

get_harvested_code_for_paper("2409.12753")
get_code_for_paper("2409.12753")
have("2409.12753")

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

curl https://syntology.ai/api/ran/2409.12753.json

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

[![Syntology run record](https://syntology.ai/api/ran/2409.12753.svg)](https://syntology.ai/paper/2409.12753)

Connect an agent — have() is free.