SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2109.01129 · ICCV · 2021

NerfingMVS: Guided Optimization of Neural Radiance Fields for Indoor Multi-view Stereo

Yongming Rao, Jiwen Lu, Jie Zhou, Wang Zhao, Shaohui Liu, Yi Wei

arXiv · PDF · Open in the Atlas

Code that ran

2 of 2 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 0 have no contract verdict; 2 failures 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/2109.01129.json · badge: /api/ran/2109.01129.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
weiyithu/nerfingmvs marked official in Papers with Code
Papers with Code archive
2 ran · 0 no recorded run
FunctionStatusWhere it lives
raw2outputs Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
weiyithu/nerfingmvs/src/run_nerf.py
code served (permissive licence) · get_code("36ddcbb8fa6be4b2")
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 tuple. 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 numpy as np

torch.manual_seed(0)

# Create small tensors with distinct sizes for different dimensions
num_rays = 2
num_samples = 4

# raw: [num_rays, num_samples along ray, 4]
# Contains RGB (first 3 channels) and density/alpha (last channel)
raw = torch.randn(num_rays, num_samples, 4)

# z_vals: [num_rays, num_samples along ray]
# Integration times along rays
z_vals = torch.linspace(0.1, 10.0, num_samples).unsqueeze(0).expand(num_rays, -1)

# rays_d: [num_rays, 3]
# Direction of each ray (should be normalized)
rays_d = torch.randn(num_rays, 3)
rays_d = rays_d / torch.norm(rays_d, dim=-1, keepdim=True)

ARGS = {
    'raw': raw,
    'z_vals': z_vals,
    'rays_d': rays_d,
    'raw_noise_std': 0.0,
    'white_bkgd': False,
    'pytest': False,
}
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 tuple — a tuple output has no array shape of its own; none is 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
render_rays Ran
ran; our contract draft was wrong, not the code
weiyithu/nerfingmvs/src/run_nerf.py
code served (permissive licence) · get_code("db6cad47ee0b8f20")
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 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
import numpy as np

torch.manual_seed(0)
np.random.seed(0)

# Define a minimal network_fn that takes (pts, viewdirs) and returns raw predictions
class MinimalNetwork(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(6, 4)  # 3 for pts + 3 for viewdirs -> 4 for RGBA
    
    def forward(self, pts, viewdirs):
        # pts shape: [num_rays, num_samples, 3]
        # viewdirs shape: [num_rays, 3] or None
        batch_size, num_samples, _ = pts.shape
        pts_flat = pts.reshape(-1, 3)
        
        if viewdirs is not None:
            viewdirs_expanded = viewdirs.unsqueeze(1).expand(batch_size, num_samples, 3)
            viewdirs_flat = viewdirs_expanded.reshape(-1, 3)
            combined = torch.cat([pts_flat, viewdirs_flat], dim=-1)
        else:
            combined = pts_flat
        
        raw = self.linear(combined)
        raw = raw.reshape(batch_size, num_samples, 4)
        return raw

network_fn = MinimalNetwork()

# Define network_query_fn that calls the network
def network_query_fn(pts, viewdirs, network_fn):
    return network_fn(pts, viewdirs)

# Create ray_batch: [num_rays, 11]
# Format: [ray_origin(3), ray_direction(3), near(1), far(1), viewdir(3)]
num_rays = 2
num_samples = 4
ray_batch = torch.randn(num_rays, 11)
ray_batch[:, 6:8] = torch.tensor([[0.5, 5.0], [0.5, 5.0]])  # near and far bounds

# Create depth_priors and depth_confidences
depth_priors = torch.ones(num_rays) * 2.0
depth_confidences = torch.ones(num_rays) * 0.1

ARGS = {
    'ray_batch': ray_batch,
    'network_fn': network_fn,
    'network_query_fn': network_query_fn,
    'N_samples': num_samples,
    'depth_priors': depth_priors,
    'depth_confidences': depth_confidences,
    'retraw': False,
    'lindisp': False,
    'perturb': 0.0,
    'white_bkgd': False,
    'raw_noise_std': 0.0,
    'pytest': False,
    'near_bound': 0.0,
    'far_bound': 1.0,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, 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

Also present in other papers' code

Of 2 samples lifted from this paper's linked repositories, 2 carry a function identity and 0 do not. Of the 2 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 2 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

NeRF [33] Ours Figure 1: Qualitative results for multi-view depth estimation on ScanNet [4]. Our method clearly surpasses leading multiview estimation methods [29, 34] by building on top of neural radiance fields [33]. While also using test-time optimization, CVD [29] suffers from inaccurate estimation of flow correspondences. NeRF [33] fails to produce accurate geometry due to the inherent shape-radiance ambiguity [61] (See Figure 3) in indoor scenes.With guided optimization, our method successfully integrates the learning-based depth priors into NeRF, significantly improving the geometry of the radiance fields.

For agents

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

get_harvested_code_for_paper("2109.01129")
get_code_for_paper("2109.01129")
have("2109.01129")

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

curl https://syntology.ai/api/ran/2109.01129.json

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

[![Syntology run record](https://syntology.ai/api/ran/2109.01129.svg)](https://syntology.ai/paper/2109.01129)

Connect an agent — have() is free.