SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2108.10869 · 2021

DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Cameras

arXiv · PDF · Open in the Atlas

Code that ran

2 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; 0 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 1 failure 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/2108.10869.json · badge: /api/ran/2108.10869.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
princeton-vl/droid-slam canonical (classifier label)
classifier basis: census
2 ran · 1 no recorded run
FunctionStatusWhere it lives
cvx_upsample Ran
ran on a synthesized input
princeton-vl/droid-slam/droid_slam/droid_net.py
code served (permissive licence) · get_code("68c92338014cace4")
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, 24, 32, 5) 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)

# Analyze the function to determine required shapes:
# - data: shape (batch, ht, wd, dim) -> permuted to (batch, dim, ht, wd)
# - mask: shape (batch, 1, 9, 8, 8, ht, wd)
# The mask has 8x8 upsampling factor embedded in dimensions 3 and 4
# F.unfold with kernel [3,3] on (batch, dim, ht, wd) produces (batch, dim*9, ht*wd)
# which is reshaped to (batch, dim, 9, 1, 1, ht, wd)
# mask * up_data performs element-wise multiplication and sum over dim 2
# Final reshape is (batch, 8*ht, 8*wd, dim)

batch = 2
ht = 3
wd = 4
dim = 5

# data: (batch, ht, wd, dim)
data = torch.randn(batch, ht, wd, dim)

# mask: (batch, 1, 9, 8, 8, ht, wd)
mask = torch.randn(batch, 1, 9, 8, 8, ht, wd)

ARGS = {
    "data": data,
    "mask": mask,
}
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, 24, 32, 5) — shape (2, 24, 32, 5) 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
upsample_disp Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
princeton-vl/droid-slam/droid_slam/droid_net.py
code served (permissive licence) · get_code("474ee7b8467698ed")
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, 32, 40) 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)

# For upsample_disp:
# - disp: shape (batch, num, ht, wd)
# - mask: shape (batch * num, ?, ht, wd) where ? comes from cvx_upsample requirements

# In cvx_upsample:
# - data: shape (batch, ht, wd, dim) after permute becomes (batch, dim, ht, wd)
# - mask: shape (batch, 1, 9, 8, 8, ht, wd)
# So mask input to cvx_upsample needs 7 dimensions

# Working backwards from upsample_disp:
# disp.view(batch * num, ht, wd, 1) -> passed as data to cvx_upsample
# mask.view(batch * num, -1, ht, wd) -> passed as mask to cvx_upsample
# 
# In cvx_upsample, mask.view(batch, 1, 9, 8, 8, ht, wd)
# So mask coming in has shape (batch * num, ?, ht, wd)
# And it needs to reshape to (batch * num, 1, 9, 8, 8, ht, wd)
# That means ? * ht * wd = 1 * 9 * 8 * 8 * ht * wd
# So ? = 9 * 8 * 8 = 576

batch = 2
num = 3
ht = 4
wd = 5
dim = 1

# disp shape: (batch, num, ht, wd)
disp = torch.randn(batch, num, ht, wd)

# mask shape: (batch * num, 576, ht, wd)
# where 576 = 9 * 8 * 8 (flattened kernel positions and upsampling grid)
mask = torch.randn(batch * num, 9 * 8 * 8, ht, wd)

ARGS = {
    "disp": disp,
    "mask": mask,
}
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, 32, 40) — shape (2, 3, 32, 40) 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
fit_scale Not yet run
unverified (harvested, no recorded run)
princeton-vl/droid-slam/droid_slam/geom/losses.py
code served (permissive licence) · get_code("fabc762095d32c21")
How it ran

no invocation recorded.

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, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 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.

FunctionAlso present inEarliest-datedOthers named
upsample_disp 1 other paper
1 later
2411.17982 (2024-11)
HI-SLAM2: Geometry-Aware Gaussian SLAM for Fast Monocular Scene Reconstruction

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 introduce DROID-SLAM, a new deep learning based SLAM system. DROID-SLAM consists of recurrent iterative updates of camera pose and pixelwise depth through a Dense Bundle Adjustment layer. DROID-SLAM is accurate, achieving large improvements over prior work, and robust, suffering from substantially fewer catastrophic failures. Despite training on monocular video, it can leverage stereo or RGB-D video to achieve improved performance at test time. The URL to our open source code is https://github.com/princeton-vl/DROID-SLAM.

For agents

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

get_harvested_code_for_paper("2108.10869")
get_code_for_paper("2108.10869")
have("2108.10869")

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

curl https://syntology.ai/api/ran/2108.10869.json

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

[![Syntology run record](https://syntology.ai/api/ran/2108.10869.svg)](https://syntology.ai/paper/2108.10869)

Connect an agent — have() is free.