Papers › LightGlue: Local Feature Matching at Light Speed

LightGlue: Local Feature Matching at Light Speed

23 Jun 2023ICCV 2023 1arXiv:2306.13643archive 2025-07-28

Philipp Lindenberger, Paul-Edouard Sarlin, Marc Pollefeys

We introduce LightGlue, a deep neural network that learns to match local features across images. We revisit multiple design decisions of SuperGlue, the state of the art in sparse matching, and derive simple but effective improvements. Cumulatively, they make LightGlue more efficient - in terms of both memory and computation, more accurate, and much easier to train. One key property is that LightGlue is adaptive to the difficulty of the problem: the inference is much faster on image pairs that are intuitively easy to match, for example because of a larger visual overlap or limited appearance change. This opens up exciting prospects for deploying deep matchers in latency-sensitive applications like 3D reconstruction. The code and trained models are publicly available at https://github.com/cvg/LightGlue.

PaperPDFConference PDFCodeCode Syntology ranMethod as code

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 40 code samples harvested from 2 repositories linked to this paper: 33 ran (of which 4 constructed an object rather than computing a result; 26 with no instrument failure: 0 honoured, 0 violated, 26 with no contract checked; 7 where Syntology's instrument failed) · 7 unverified. Where Syntology's instrument failed: 2 where our contract draft was wrong; 5 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 30 samples from 1 repository, 26 ran (21 with no instrument failure; 5 where Syntology's instrument failed); community (archive-listed): 10 samples from 1 repository, 7 ran (5 with no instrument failure; 2 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written between 2026-09-22 and 2026-09-26 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

cvg/lightglue officialmentioned in papermentioned on GitHubpytorchApache-2.0Syntology: 30 samples, 26 ran (5 instrument-failed) · 4 unverified report
fabio-sim/LightGlue-ONNX mentioned on GitHubpytorchSyntology: 10 samples, 7 ran (2 instrument-failed) · 3 unverified report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

40 samples harvested: 33 ran (of which 4 constructed an object rather than computing a result; 26 with no instrument failure: 0 honoured, 0 violated, 26 with no contract checked; 7 where Syntology's instrument failed) · 7 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

2ran · our draft was wrong
5ran; the contract check could not run
26ran
7unverified

Licence: 0 of the 40 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 2 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. For 4 samples that ran, the graph records a signature this page does not show: for a class the fixture built without calling, the harness called the object with inputs it made up itself, so the recorded signature is the harness's probe, not the fixture's arguments. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

cvg/lightglue official repository 30 samples, 26 ran (5 instrument-failed) · 4 unverified
Attention in 344 papers
census: 13 implementations from 12 papers, 10 distinct outputs across 4 buckets, one shared input per bucket this sample: bucket 2 of 4, arg 1: rank 3, kind float, dtype float32 · arg 2: rank 3, kind float, dtype float32 · arg 3: rank 3, kind float, dtype float32
lightglue/lightglue.py ran · metamorphic tier: invariant fingerprinted Apache-2.0 (permissive) · a08f05b65312af52 · report fixture: signature q:Tensor/3/float, k:Tensor/3/float, v:Tensor/3/float, mask:NoneType/0/? · returned Tensor · level recorded 2026-09-22
LearnableFourierPositionalEncoding lightglue/lightglue.py ran · metamorphic tier: invariant Apache-2.0 (permissive) · 417b02cca4cc3fd2 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-22
MatchAssignment lightglue/lightglue.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · b47a54dd8d0e3728 · report fixture: signature desc0:Tensor/3/float, desc1:Tensor/3/float · returned tuple · level recorded 2026-09-22
SelfBlock lightglue/lightglue.py ran Apache-2.0 (permissive) · 2caf26511a6409ff · report fixture: signature x:Tensor/3/float, encoding:Tensor/3/float, mask:NoneType/0/? · returned Tensor · level recorded 2026-09-24
TokenConfidence lightglue/lightglue.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · 57de8d63639644e2 · report fixture: signature desc0:Tensor/2/float, desc1:Tensor/2/float · returned tuple · level recorded 2026-09-22
apply_cached_rotary_emb lightglue/lightglue.py ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · 127ac20ae2475310 · report fixture: signature freqs:tuple/1/Tensor, t:Tensor/2/float · returned Tensor · level recorded 2026-09-22
batch_to_device in 18 papers lightglue/utils.py ran Apache-2.0 (permissive) · 80f11116ef08b1e5 · report fixture: signature batch:dict/0/?, device:str/0/str, non_blocking:bool/0/bool · returned dict · level recorded 2026-09-26
cm_BlRdGn lightglue/viz2d.py ran Apache-2.0 (permissive) · 8dd996cdd1db911e · report fixture: signature x_:ndarray/2/float · returned ndarray · level recorded 2026-09-26
cm_RdGn lightglue/viz2d.py ran fingerprinted Apache-2.0 (permissive) · 736e2b86a021d585 · report fixture: signature x:ndarray/1/float · returned ndarray · level recorded 2026-09-26
cm_prune lightglue/viz2d.py ran fingerprinted Apache-2.0 (permissive) · 305ee29b4e31eae6 · report fixture: signature x_:ndarray/1/float · returned ndarray · level recorded 2026-09-26
filter_dog_point lightglue/sift.py ran Apache-2.0 (permissive) · 8d37ad51dbe2ac79 · report fixture: signature points:ndarray/2/float, scales:ndarray/1/float, angles:ndarray/1/float, image_shape:tuple/1/int, nms_radius:int/0/int, scores:ndarray/1/float · returned ndarray · level recorded 2026-09-26
filter_matches lightglue/lightglue.py ran; the contract check could not run (RAISES) fingerprinted Apache-2.0 (permissive) · 43dbd6f5401d29db · report fixture: signature scores:Tensor/3/float, th:float/0/float · returned tuple · level recorded 2026-09-22
get_conv in 21 papers lightglue/aliked.py ran Apache-2.0 (permissive) · c862b64639150d6c · report fixture: signature inplanes:int/0/int, planes:int/0/int, kernel_size:int/0/int, stride:int/0/int, padding:int/0/int, bias:bool/0/bool, conv_type:str/0/str, mask:bool/0/bool · returned DeformableConv2d · level recorded 2026-09-26
get_patches in 9 papers lightglue/aliked.py ran Apache-2.0 (permissive) · c909ca701cba5a2b · report fixture: signature tensor:Tensor/3/float, required_corners:Tensor/2/float, ps:int/0/int · returned Tensor · level recorded 2026-09-26
map_tensor lightglue/utils.py ran Apache-2.0 (permissive) · 6cf207b8475b5658 · report fixture: signature input_:dict/0/?, func:callable/0/? · returned dict · level recorded 2026-09-26
measure in 9 papers benchmark.py ran Apache-2.0 (permissive) · 2bd5dd5a07a0f3be · report fixture: signature matcher:callable/0/?, data:Tensor/2/float, device:device/0/?, r:int/0/int · returned dict · level recorded 2026-09-26
pad_to_length in 18 papers lightglue/lightglue.py ran; the contract check could not run (RAISES) fingerprinted Apache-2.0 (permissive) · 8751587f16cda36c · report fixture: signature x:Tensor/3/float, length:int/0/int · returned tuple · level recorded 2026-09-22
rbd lightglue/utils.py ran Apache-2.0 (permissive) · 1c2f9a87f4978298 · report fixture: signature data:dict/0/? · returned dict · level recorded 2026-09-26
rotate_half in 424 papers
census: 24 implementations from 96 papers, 4 distinct outputs across 2 buckets, one shared input per bucket this sample: bucket 1 of 2, arg 1: rank 2, kind float, dtype float32
lightglue/lightglue.py ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · a3ec43c12656acbb · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-22
run_opencv_sift lightglue/sift.py ran Apache-2.0 (permissive) · 29d472991fc656bb · report fixture: signature features:SIFT/0/?, image:ndarray/2/int · returned tuple · level recorded 2026-09-26
sample_descriptors lightglue/superpoint.py ran Apache-2.0 (permissive) · 04a4d6560cc90824 · report fixture: signature keypoints:Tensor/3/float, descriptors:Tensor/4/float, s:int/0/int · returned Tensor · level recorded 2026-09-26
sift_to_rootsift lightglue/sift.py ran fingerprinted Apache-2.0 (permissive) · 971d8cc47145eba6 · report fixture: signature x:Tensor/2/float, eps:float/0/float · returned Tensor · level recorded 2026-09-26
sigmoid_log_double_softmax lightglue/lightglue.py ran; the contract check could not run (RAISES) Apache-2.0 (permissive) · 608069c6c4b78f43 · report fixture: signature sim:Tensor/3/float, z0:Tensor/3/float, z1:Tensor/3/float · returned Tensor · level recorded 2026-09-22
simple_nms lightglue/aliked.py ran fingerprinted Apache-2.0 (permissive) · 5c5360d394e7680e · report fixture: signature scores:Tensor/4/float, nms_radius:int/0/int · returned Tensor · level recorded 2026-09-26
simple_nms lightglue/superpoint.py ran fingerprinted Apache-2.0 (permissive) · 30f2d08994bffec3 · report fixture: signature scores:Tensor/4/float, nms_radius:int/0/int · returned Tensor · level recorded 2026-09-26
top_k_keypoints lightglue/superpoint.py ran Apache-2.0 (permissive) · 719478e41beaad68 · report fixture: signature keypoints:Tensor/2/float, scores:Tensor/1/float, k:int/0/int · returned tuple · level recorded 2026-09-24
CrossBlock lightglue/lightglue.py unverified Apache-2.0 (permissive) · ee69084527f306a8 · report
LightGlue lightglue/lightglue.py unverified Apache-2.0 (permissive) · e3b3b3c0e63763bd · report
TransformerLayer in 34 papers lightglue/lightglue.py unverified Apache-2.0 (permissive) · 0a5c71b088578afa · report
normalize_keypoints in 10 papers lightglue/lightglue.py unverified Apache-2.0 (permissive) · 996031e8e332cf27 · report
fabio-sim/LightGlue-ONNX community (archive-listed) 10 samples, 7 ran (2 instrument-failed) · 3 unverified
CrossBlock lightglue_dynamo/models/lightglue.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · d8cc2fdd4e6e8b77 · report fixture: returned CrossBlock · level recorded 2026-09-22
LearnableFourierPositionalEncoding lightglue_dynamo/models/lightglue.py ran · metamorphic tier: deterministic fingerprinted Apache-2.0 (permissive) · b7828f594fbf256c · report fixture: returned LearnableFourierPositionalEncoding · level recorded 2026-09-22
MatchAssignment lightglue_dynamo/models/lightglue.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · d48ee5cc4cbd32b3 · report fixture: returned MatchAssignment · level recorded 2026-09-22
SelfBlock lightglue_dynamo/models/lightglue.py ran Apache-2.0 (permissive) · be7cd4498eab4bc0 · report fixture: signature x:Tensor/3/float, encoding:tuple/1/Tensor · returned Tensor · level recorded 2026-09-24
TokenConfidence lightglue_dynamo/models/lightglue.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · 1c0939b0051d6ca5 · report fixture: returned TokenConfidence · level recorded 2026-09-22
filter_matches lightglue_dynamo/models/lightglue.py ran; the contract check could not run (RAISES) · metamorphic tier: deterministic fingerprinted Apache-2.0 (permissive) · f485836d645eb208 · report fixture: signature scores:Tensor/3/float, threshold:float/0/float · returned tuple · level recorded 2026-09-22
multi_head_attention in 24 papers lightglue_dynamo/models/lightglue.py ran; the contract check could not run (RAISES) fingerprinted Apache-2.0 (permissive) · 7350588d723a13e5 · report fixture: signature q:Tensor/3/float, k:Tensor/3/float, v:Tensor/3/float, num_heads:int/0/int · returned Tensor · level recorded 2026-09-22
LightGlue lightglue_dynamo/models/lightglue.py unverified Apache-2.0 (permissive) · a7c63934348a934f · report
TransformerLayer in 34 papers lightglue_dynamo/models/lightglue.py unverified Apache-2.0 (permissive) · ab4f90d2792e9808 · report
sigmoid_log_double_softmax lightglue_dynamo/models/lightglue.py unverified Apache-2.0 (permissive) · 7eb939455e9a0658 · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it. Siblings are assemblies lifted from this paper in the same batch for the same method: alternative readings kept side by side; none replaces or supersedes another.

2 assemblies on this paper: 2 harvested, 0 generated, listed in 1 groups (siblings lifted in one batch form one group).

2 sibling assemblies for LightGlue (harvested), lifted in the same batch on 2026-09-09: alternative readings kept side by side; none replaces or supersedes another.

harvested Assembly for LightGlue (graph method: LightGlue). 10 bricks; individually on a fixture: 7 ran (of which 4 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 0 violated, 5 with no contract checked; 2 where Syntology's instrument failed) · 3 unverified. The assembly has not been run end-to-end.

one of 2 siblings

  1. LightGlue entry · level 0 · unverified · a7c63934348a934f
  2. LearnableFourierPositionalEncoding model · level 1 · ran · b7828f594fbf256c
  3. MatchAssignment model · level 1 · ran · d48ee5cc4cbd32b3
  4. TokenConfidence model · level 1 · ran · 1c0939b0051d6ca5
  5. TransformerLayer model · level 0 · unverified · ab4f90d2792e9808
  6. filter_matches helper · level 1 · ran; the contract check could not run (RAISES) · f485836d645eb208
  7. sigmoid_log_double_softmax helper · level 0 · unverified · 7eb939455e9a0658
  8. CrossBlock model · level 1 · ran · d8cc2fdd4e6e8b77
  9. SelfBlock model · level 1 · ran · be7cd4498eab4bc0
  10. multi_head_attention helper · level 1 · ran; the contract check could not run (RAISES) · 7350588d723a13e5

harvested Assembly for LightGlue (graph method: LightGlue). 14 bricks; individually on a fixture: 10 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 0 violated, 5 with no contract checked; 5 where Syntology's instrument failed) · 4 unverified. The assembly has not been run end-to-end.

one of 2 siblings

  1. LightGlue entry · level 0 · unverified · e3b3b3c0e63763bd
  2. LearnableFourierPositionalEncoding helper · level 1 · ran · 417b02cca4cc3fd2
  3. MatchAssignment helper · level 1 · ran · b47a54dd8d0e3728
  4. TokenConfidence helper · level 1 · ran · 57de8d63639644e2
  5. TransformerLayer helper · level 0 · unverified · 0a5c71b088578afa
  6. filter_matches helper · level 1 · ran; the contract check could not run (RAISES) · 43dbd6f5401d29db
  7. normalize_keypoints helper · level 0 · unverified · 996031e8e332cf27
  8. pad_to_length helper · level 1 · ran; the contract check could not run (RAISES) · 8751587f16cda36c
  9. sigmoid_log_double_softmax helper · level 1 · ran; the contract check could not run (RAISES) · 608069c6c4b78f43
  10. CrossBlock helper · level 0 · unverified · ee69084527f306a8
  11. SelfBlock helper · level 1 · ran · 2caf26511a6409ff
  12. Attention helper · level 1 · ran · a08f05b65312af52
  13. apply_cached_rotary_emb helper · level 1 · ran · our draft was wrong · 127ac20ae2475310
  14. rotate_half helper · level 1 · ran · our draft was wrong · a3ec43c12656acbb

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2306.13643") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

3D ReconstructionCamera Pose EstimationHomography EstimationImage MatchingKey Point MatchingPose EstimationVisual Localization

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Matching ZEB LightGlue Mean AUC@5° 31.7 #9 of 10 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections