Papers › Revealing the Dark Secrets of Masked Image Modeling

Revealing the Dark Secrets of Masked Image Modeling

26 May 2022CVPR 2023 1arXiv:2205.13543archive 2025-07-28

Zhenda Xie, Zigang Geng, Jingcheng Hu, Zheng Zhang, Han Hu, Yue Cao

Masked image modeling (MIM) as pre-training is shown to be effective for numerous vision downstream tasks, but how and where MIM works remain unclear. In this paper, we compare MIM with the long-dominant supervised pre-trained models from two perspectives, the visualizations and the experiments, to uncover their key representational differences. From the visualizations, we find that MIM brings locality inductive bias to all layers of the trained models, but supervised models tend to focus locally at lower layers but more globally at higher layers. That may be the reason why MIM helps Vision Transformers that have a very large receptive field to optimize. Using MIM, the model can maintain a large diversity on attention heads in all layers. But for supervised models, the diversity on attention heads almost disappears from the last three layers and less diversity harms the fine-tuning performance. From the experiments, we find that MIM models can perform significantly better on geometric and motion tasks with weak semantics or fine-grained classification tasks, than their supervised counterparts. Without bells and whistles, a standard MIM pre-trained SwinV2-L could achieve state-of-the-art performance on pose estimation (78.9 AP on COCO test-dev and 78.0 AP on CrowdPose), depth estimation (0.287 RMSE on NYUv2 and 1.966 RMSE on KITTI), and video object tracking (70.7 SUC on LaSOT). For the semantic understanding datasets where the categories are sufficiently covered by the supervised pre-training, MIM models can still achieve highly competitive transfer performance. With a deeper understanding of MIM, we hope that our work can inspire new and solid research in this direction.

PaperPDFConference PDFCodeCode Syntology ran

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 5 code samples harvested from 1 repository linked to this paper: 5 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 1 violated, 2 with no contract checked; 2 where Syntology's instrument failed) · 0 unverified. Where Syntology's instrument failed: 1 where our contract draft was wrong; 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 5 samples from 1 repository, 5 ran (3 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 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 1 of the 5 samples that ran has no recorded date.

SwinTransformer/MIM-Depth-Estimation officialmentioned on GitHubpytorchMITSyntology: 5 samples, 5 ran (2 instrument-failed) · 0 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

5 samples harvested: 5 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 1 violated, 2 with no contract checked; 2 where Syntology's instrument failed) · 0 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.

1ran · violated contract
1ran · our draft was wrong
1ran; the contract check could not run
2ran

Licence: 0 of the 5 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 SwinTransformer/MIM-Depth-Estimation. “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”. “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.

SwinTransformer/MIM-Depth-Estimation official repository 5 samples, 5 ran (2 instrument-failed) · 0 unverified
get_dataset in 285 papers dataset/base_dataset.py ran MIT (permissive) · 1828822ff8e88962 · report fixture: signature dataset_name:str/0/str · returned MockDataset · level recorded 2026-09-24
get_num_layer_for_swin models/optimizer.py ran MIT (permissive) · 6907f72868217c02 · report fixture: signature var_name:str/0/str, num_max_layer:int/0/int, layers_per_stage:list/1/int · returned int · level recorded 2026-09-24
str2bool in 533 papers configs/base_options.py ran · violated contract MIT (permissive) · 7c508037b40522af · report fixture: signature v:str/0/str · returned bool · level recorded, date not recorded
window_partition in 377 papers
census: 13 implementations from 17 papers, 5 distinct outputs across 3 buckets, one shared input per bucket
models/swin_transformer_v2.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted MIT (permissive) · f9fd6241d935f07b · report fixture: signature x:Tensor/4/float, window_size:int/0/int · returned Tensor · level recorded 2026-09-22
window_reverse in 296 papers models/swin_transformer_v2.py ran · our draft was wrong MIT (permissive) · fb32094c6dbece71 · report fixture: signature windows:Tensor/4/float, window_size:int/0/int, H:int/0/int, W:int/0/int · returned Tensor · level recorded 2026-09-22

Tasks

Depth EstimationDiversityInductive BiasMonocular Depth EstimationObject TrackingPose EstimationVideo Object TrackingVisual Object Tracking

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Depth Estimation NYU-Depth V2 SwinV2-L 1K-MIM RMS 0.287 #3 of 17 Archive leaderboard report
Depth Estimation NYU-Depth V2 SwinV2-B 1K-MIM RMS 0.304 #5 of 17 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-L 1K-MIM Delta < 1.25 0.977 #17 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-L 1K-MIM Delta < 1.25^2 0.998 #17 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-L 1K-MIM Delta < 1.25^3 1.000 #17 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-L 1K-MIM RMSE 1.966 #17 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-L 1K-MIM RMSE log 0.075 #17 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-L 1K-MIM Sq Rel 0.139 #17 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-L 1K-MIM absolute relative error 0.050 #17 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-B 1K-MIM Delta < 1.25 0.976 #25 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-B 1K-MIM Delta < 1.25^2 0.998 #25 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-B 1K-MIM Delta < 1.25^3 0.999 #25 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-B 1K-MIM RMSE 2.050 #25 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-B 1K-MIM RMSE log 0.078 #25 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-B 1K-MIM Sq Rel 0.148 #25 of 79 Archive leaderboard report
Monocular Depth Estimation KITTI Eigen split SwinV2-B 1K-MIM absolute relative error 0.052 #25 of 79 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 SwinV2-L 1K-MIM Delta < 1.25 0.949 #29 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 SwinV2-L 1K-MIM Delta < 1.25^2 0.994 #29 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 SwinV2-L 1K-MIM Delta < 1.25^3 0.999 #29 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 SwinV2-L 1K-MIM RMSE 0.287 #29 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 SwinV2-L 1K-MIM absolute relative error 0.083 #29 of 85 Archive leaderboard report
Monocular Depth Estimation NYU-Depth V2 SwinV2-L 1K-MIM log 10 0.035 #29 of 85 Archive leaderboard report
Pose Estimation COCO test-dev SwinV2-L 1K-MIM AP 77.2 #11 of 47 Archive leaderboard report
Pose Estimation COCO test-dev SwinV2-B 1K-MIM AP 76.7 #14 of 47 Archive leaderboard report
Pose Estimation CrowdPose SwinV2-L 1K-MIM AP 75.5 #4 of 12 Archive leaderboard report
Pose Estimation CrowdPose SwinV2-B 1K-MIM AP 74.9 #5 of 12 Archive leaderboard report
Visual Object Tracking GOT-10k SwinV2-L 1K-MIM Average Overlap 72.9 #24 of 42 Archive leaderboard report
Visual Object Tracking GOT-10k SwinV2-B 1K-MIM Average Overlap 70.8 #26 of 42 Archive leaderboard report
Visual Object Tracking LaSOT SwinV2-L 1K-MIM AUC 70.7 #26 of 46 Archive leaderboard report
Visual Object Tracking LaSOT SwinV2-B 1K-MIM AUC 70 #31 of 46 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.

Methods

MIM

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