SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2304.12461 · CVPR · 2023

TensoIR: Tensorial Inverse Rendering

Hao Su, Xiaoshuai Zhang, Xiaowei Zhou, Sai Bi, Zexiang Xu, Haian Jin, Isabella Liu, Peijia Xu, Songfang Han

arXiv · PDF · Open in the Atlas

Code that ran

4 of 5 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; 0 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 2 failures of our instrument; 1 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/2304.12461.json · badge: /api/ran/2304.12461.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
Haian-Jin/TensoIR canonical (classifier label)
classifier basis: census
4 ran · 1 no recorded run
FunctionStatusWhere it lives
eval_sh Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Haian-Jin/TensoIR/models/sh.py
pointer only (licence: MIT); the stored copy get_code serves, from pcl3dv/ov-nerf, is pointer only (licence: NONE) · get_code("db95ecdf5d4abe6b")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from pcl3dv/ov-nerf, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

eval_sh_bases Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Haian-Jin/TensoIR/models/sh.py
code served (permissive licence) · get_code("caa9e209a9083a0f")
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 2, dtype family float, shape (2, 9) 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 eval_sh_bases takes:
# - deg: int, SH max degree (0-4)
# - dirs: torch.Tensor of shape (..., 3) representing unit directions

# Create a small batch of unit direction vectors
# Shape: (2, 3) - batch of 2 directions, each with 3 components (x, y, z)
dirs = torch.tensor([
    [1.0, 0.0, 0.0],
    [0.0, 1.0, 0.0]
], dtype=torch.float32)

# Normalize to ensure they are unit directions
dirs = dirs / torch.norm(dirs, dim=-1, keepdim=True)

# Test with degree 2 (a middle value)
deg = 2

ARGS = {
    "deg": deg,
    "dirs": dirs
}
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 2, dtype family float, shape (2, 9) — shape (2, 9) 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
grid_sample Ran
ran on a synthesized input
Haian-Jin/TensoIR/models/relight_utils.py
code served (permissive licence) · get_code("ababd7f4b87e7d34")
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, 3, 3, 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)

# Create a small image tensor: (N, C, IH, IW)
# N=2 (batch), C=3 (channels), IH=4 (image height), IW=5 (image width)
image = torch.randn(2, 3, 4, 5)

# Create optical flow tensor: (N, H, W, 2)
# N=2 (batch), H=3 (output height), W=4 (output width), 2 (x, y coordinates)
# Optical flow values should be in range [-1, 1] for normalized grid sampling
optical = torch.rand(2, 3, 4, 2) * 2 - 1  # Values in [-1, 1]

ARGS = {
    "image": image,
    "optical": optical,
}
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, 3, 3, 4) — shape (2, 3, 3, 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
safe_l2_normalize Ran
ran on a synthesized input
Haian-Jin/TensoIR/models/relight_utils.py
code served (permissive licence) · get_code("a53a50041760e47c")
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.functional as F

torch.manual_seed(0)

# Create a small tensor for normalization
x = torch.randn(2, 4)

# The function takes x, dim (optional), and eps (optional)
# dim can be None, an int, or a tuple of ints
# eps is a float for numerical stability

ARGS = {
    "x": x,
    "dim": None,
    "eps": 1e-06
}
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
GGX_specular Not yet run
unverified (harvested, no recorded run)
Haian-Jin/TensoIR/models/relight_utils.py
code served (permissive licence) · get_code("677a27c444c9af83")
How it ran

no invocation recorded.

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 2 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.

FunctionAlso present inEarliest-datedOthers named
eval_sh_bases 15 other papers
4 dated earlier, 11 later
2301.02238 (2023-01)
HyperReel: High-Fidelity 6-DoF Video with Ray-Conditioned Sampling
2301.09632 (2023-01)
HexPlane: A Fast Representation for Dynamic Scenes
2302.01226 (2023-02)
Factor Fields: A Unified Framework for Neural Fields and Beyond
and 12 more not named here
eval_sh 14 other papers
3 dated earlier, 11 later
2301.09632 (2023-01)
HexPlane: A Fast Representation for Dynamic Scenes
2302.01226 (2023-02)
Factor Fields: A Unified Framework for Neural Fields and Beyond
2303.03808 (2023-03)
Multiscale Tensor Decomposition and Rendering Equation Encoding for View Synthesis
and 11 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 TensoIR, a novel inverse rendering approach based on tensor factorization and neural fields. Unlike previous works that use purely MLP-based neural fields, thus suffering from low capacity and high computation costs, we extend TensoRF, a state-of-the-art approach for radiance field modeling, to estimate scene geometry, surface reflectance, and environment illumination from multiview images captured under unknown lighting conditions. Our approach jointly achieves radiance field reconstruction and physically-based model estimation, leading to photorealistic novel view synthesis and relighting results. Benefiting from the efficiency and extensibility of the TensoRFbased representation, our method can accurately model secondary shading effects (like shadows and indirect lighting) and generally support input images captured under single or multiple unknown lighting conditions. The low-rank tensor representation allows us to not only achieve fast and compact reconstruction but also better exploit shared information under an arbitrary number of capturing lighting conditions. We demonstrate the superiority of our method to baseline methods qualitatively and quantitatively on various challenging synthetic and real-world scenes.

For agents

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

get_harvested_code_for_paper("2304.12461")
get_code_for_paper("2304.12461")
have("2304.12461")

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

curl https://syntology.ai/api/ran/2304.12461.json

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

[![Syntology run record](https://syntology.ai/api/ran/2304.12461.svg)](https://syntology.ai/paper/2304.12461)

Connect an agent — have() is free.