Papers › Masked Autoencoders that Listen

Masked Autoencoders that Listen

13 Jul 2022arXiv:2207.06405archive 2025-07-28

Po-Yao Huang, Hu Xu, Juncheng Li, Alexei Baevski, Michael Auli, Wojciech Galuba, Florian Metze, Christoph Feichtenhofer

This paper studies a simple extension of image-based Masked Autoencoders (MAE) to self-supervised representation learning from audio spectrograms. Following the Transformer encoder-decoder design in MAE, our Audio-MAE first encodes audio spectrogram patches with a high masking ratio, feeding only the non-masked tokens through encoder layers. The decoder then re-orders and decodes the encoded context padded with mask tokens, in order to reconstruct the input spectrogram. We find it beneficial to incorporate local window attention in the decoder, as audio spectrograms are highly correlated in local time and frequency bands. We then fine-tune the encoder with a lower masking ratio on target datasets. Empirically, Audio-MAE sets new state-of-the-art performance on six audio and speech classification tasks, outperforming other recent models that use external supervised pre-training. The code and models will be at https://github.com/facebookresearch/AudioMAE.

PaperPDFCodeCode 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 32 code samples harvested from 4 repositories linked to this paper: 17 ran (of which 12 constructed an object rather than computing a result; 16 with no instrument failure: 1 honoured, 1 violated, 14 with no contract checked; 1 where Syntology's instrument failed) · 15 unverified. Where Syntology's instrument failed: 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 4 samples from 1 repository, 1 ran (1 with no instrument failure; 0 where Syntology's instrument failed); community (archive-listed): 28 samples from 3 repositories, 16 ran (15 with no instrument failure; 1 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 on 2026-09-22 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

facebookresearch/audiomae officialmentioned in papermentioned on GitHubpytorchNOASSERTIONSyntology: 4 samples, 1 ran (0 instrument-failed) · 3 unverified report
eml-eda/tle-supervised mentioned on GitHubpytorchSyntology: 6 samples, 5 ran (1 instrument-failed) · 1 unverified report
facebookresearch/multimodal mentioned on GitHubpytorchSyntology: 11 samples, 2 ran (0 instrument-failed) · 9 unverified report
rishikksh20/AudioMAE-pytorch mentioned on GitHubpytorchSyntology: 11 samples, 9 ran (0 instrument-failed) · 2 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

32 samples harvested: 17 ran (of which 12 constructed an object rather than computing a result; 16 with no instrument failure: 1 honoured, 1 violated, 14 with no contract checked; 1 where Syntology's instrument failed) · 15 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 · honoured contract
1ran · violated contract
1ran; the contract check could not run
14ran
15unverified

Licence: 10 of the 32 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 4 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 10 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.

facebookresearch/audiomae official repository 4 samples, 1 ran (0 instrument-failed) · 3 unverified
PatchEmbed_org models_mae.py ran · metamorphic tier: deterministic licence not identified · pointer only · aaa7fea64dd2b57c · report fixture: returned PatchEmbed_org · level recorded 2026-09-22
MaskedAutoencoderViT in 14 papers models_mae.py unverified licence not identified · pointer only · 28128850f3b7974d · report
PatchEmbed_new models_mae.py unverified licence not identified · pointer only · d28d7b155c97688a · report
get_2d_sincos_pos_embed_flexible in 6 papers models_mae.py unverified licence not identified · pointer only · 728825dfce4fad97 · report
eml-eda/tle-supervised community (archive-listed) 6 samples, 5 ran (1 instrument-failed) · 1 unverified
FeedForward in 198 papers models/models_audio_mae.py ran · metamorphic tier: deterministic no licence file found · pointer only · a78cda64d88bd1f2 · report fixture: returned FeedForward · level recorded 2026-09-22
SwinBlock models/models_audio_mae.py ran · metamorphic tier: deterministic no licence file found · pointer only · 0caaefeb6252c9e5 · report fixture: returned SwinBlock · level recorded 2026-09-22
Transformer in 173 papers models/models_audio_mae.py ran · metamorphic tier: deterministic no licence file found · pointer only · 8ae2d3c8074c9653 · report fixture: returned Transformer · level recorded 2026-09-22
WindowAttention in 50 papers models/models_audio_mae.py ran · metamorphic tier: deterministic no licence file found · pointer only · 6322ff428c4c429a · report fixture: returned WindowAttention · level recorded 2026-09-22
get_2d_sincos_pos_embed in 170 papers models/models_audio_mae.py ran; the contract check could not run (RAISES) no licence file found · pointer only · 11b4cbd5cc85097d · report fixture: signature embed_dim:int/0/int, grid_sizes:tuple/1/int, cls_token:bool/0/bool · returned ndarray · level recorded 2026-09-22
AudioMaskedAutoencoderViT models/models_audio_mae.py unverified no licence file found · pointer only · b9ab992d4f387299 · report
facebookresearch/multimodal community (archive-listed) 11 samples, 2 ran (0 instrument-failed) · 9 unverified
MLP in 375 papers torchmultimodal/models/masked_auto_encoder/model.py ran · metamorphic tier: invariant BSD-3-Clause (permissive) · 084f9fe01c1b3cf0 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-22
MultiHeadSelfAttention in 17 papers torchmultimodal/models/masked_auto_encoder/model.py ran · metamorphic tier: invariant fingerprinted BSD-3-Clause (permissive) · 354dccecc7d137d8 · report fixture: signature query:Tensor/3/float, attn_mask:NoneType/0/?, is_causal:bool/0/bool · returned Tensor · level recorded 2026-09-22
MAEOutput torchmultimodal/models/masked_auto_encoder/model.py unverified BSD-3-Clause (permissive) · a848037ef9265b59 · report
MaskedAutoEncoder torchmultimodal/models/masked_auto_encoder/model.py unverified BSD-3-Clause (permissive) · 680afbcfbca11fb9 · report
SwinTransformer in 15 papers torchmultimodal/models/masked_auto_encoder/model.py unverified BSD-3-Clause (permissive) · fef623fdfcfbe69b · report
SwinTransformerBlock in 35 papers torchmultimodal/models/masked_auto_encoder/model.py unverified BSD-3-Clause (permissive) · 599cc34da67f0356 · report
TransformerEncoder in 96 papers torchmultimodal/models/masked_auto_encoder/model.py unverified BSD-3-Clause (permissive) · fd67d2fafb66524c · report
TransformerEncoderLayer in 90 papers torchmultimodal/models/masked_auto_encoder/model.py unverified BSD-3-Clause (permissive) · 7a00feceab29852a · report
TransformerOutput in 5 papers torchmultimodal/models/masked_auto_encoder/model.py unverified BSD-3-Clause (permissive) · f7d645cfcd41319b · report
WindowMultiHeadAttention torchmultimodal/models/masked_auto_encoder/model.py unverified BSD-3-Clause (permissive) · bcffcc5e0273f4c1 · report
audio_mae torchmultimodal/models/masked_auto_encoder/model.py unverified BSD-3-Clause (permissive) · 8efb3224591920ef · report
rishikksh20/AudioMAE-pytorch community (archive-listed) 11 samples, 9 ran (0 instrument-failed) · 2 unverified
Attention in 344 papers
census: 13 implementations from 12 papers, 10 distinct outputs across 4 buckets, one shared input per bucket
audio_mae.py ran · metamorphic tier: deterministic MIT (permissive) · 3bd6d4930cb3f2f8 · report fixture: returned Attention · level recorded 2026-09-22
CyclicShift audio_mae.py ran · metamorphic tier: deterministic fingerprinted MIT (permissive) · b20098ca471958b8 · report fixture: returned CyclicShift · level recorded 2026-09-22
PatchEmbed in 123 papers
census: 9 implementations from 9 papers, 6 distinct outputs on one shared input
audio_mae.py ran · metamorphic tier: deterministic MIT (permissive) · c691b3b94f34ae10 · report fixture: returned PatchEmbed · level recorded 2026-09-22
Residual in 24 papers audio_mae.py ran · metamorphic tier: deterministic fingerprinted MIT (permissive) · c824fe1a650a3f66 · report fixture: returned Residual · level recorded 2026-09-22
SwinBlock audio_mae.py ran · metamorphic tier: deterministic MIT (permissive) · 22e4ee062fba7a5e · report fixture: returned SwinBlock · level recorded 2026-09-22
Transformer in 173 papers audio_mae.py ran · metamorphic tier: deterministic MIT (permissive) · 54a9926f9e0936b8 · report fixture: returned Transformer · level recorded 2026-09-22
WindowAttention in 50 papers audio_mae.py ran · metamorphic tier: deterministic MIT (permissive) · d7227307e8862fc8 · report fixture: returned WindowAttention · level recorded 2026-09-22
create_mask in 35 papers audio_mae.py ran · violated contract MIT (permissive) · 5b1dded6034b41e3 · report fixture: signature window_size:int/0/int, displacement:int/0/int, upper_lower:bool/0/bool, left_right:bool/0/bool · returned Tensor · level recorded 2026-09-22
get_relative_distances in 6 papers audio_mae.py ran · honoured contract MIT (permissive) · ea931f8eaca3c48c · report fixture: signature window_size:int/0/int · returned Tensor · level recorded 2026-09-22
AudioMaskedAutoencoderViT audio_mae.py unverified MIT (permissive) · c263573424bacb51 · report
get_2d_sincos_pos_embed in 170 papers audio_mae.py unverified MIT (permissive) · 4dafab91858a28d2 · 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.

4 assemblies on this paper: 4 harvested, 0 generated, listed in 1 groups (siblings lifted in one batch form one group).

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

harvested Assembly for Audio-MAE (graph method: Audio-MAE). 15 bricks; individually on a fixture: 11 ran (of which 9 constructed an object rather than computing a result; 11 with no instrument failure: 1 honoured, 1 violated, 9 with no contract checked; 0 where Syntology's instrument failed) · 4 unverified. The assembly has not been run end-to-end.

one of 4 siblings

  1. AudioMaskedAutoencoderViT entry · level 0 · unverified · c263573424bacb51
  2. PatchEmbed model · level 1 · ran · c691b3b94f34ae10
  3. SwinBlock helper · level 1 · ran · 22e4ee062fba7a5e
  4. Transformer model · level 1 · ran · 54a9926f9e0936b8
  5. get_2d_sincos_pos_embed helper · level 0 · unverified · 4dafab91858a28d2
  6. FeedForward model · level 1 · ran · c17d96c99b861de1
  7. PreNorm helper · level 1 · ran · 005532cda6b215dd
  8. Residual helper · level 1 · ran · c824fe1a650a3f66
  9. WindowAttention helper · level 1 · ran · d7227307e8862fc8
  10. Attention model · level 1 · ran · 3bd6d4930cb3f2f8
  11. get_2d_sincos_pos_embed_from_grid helper · level 0 · unverified · f10004e059714d42
  12. CyclicShift helper · level 1 · ran · b20098ca471958b8
  13. create_mask helper · level 1 · ran · violated contract · 5b1dded6034b41e3
  14. get_relative_distances helper · level 1 · ran · honoured contract · ea931f8eaca3c48c
  15. get_1d_sincos_pos_embed_from_grid helper · level 0 · unverified · 12035a2f77d8016c

harvested Assembly for Audio-MAE (graph method: Audio-MAE). 15 bricks; individually on a fixture: 13 ran (of which 9 constructed an object rather than computing a result; 11 with no instrument failure: 1 honoured, 1 violated, 9 with no contract checked; 2 where Syntology's instrument failed) · 2 unverified. The assembly has not been run end-to-end.

one of 4 siblings

  1. AudioMaskedAutoencoderViT entry · level 0 · unverified · b9ab992d4f387299
  2. PatchEmbed model · level 1 · ran · c691b3b94f34ae10
  3. SwinBlock helper · level 1 · ran · 0caaefeb6252c9e5
  4. Transformer model · level 1 · ran · 8ae2d3c8074c9653
  5. get_2d_sincos_pos_embed helper · level 1 · ran; the contract check could not run (RAISES) · 11b4cbd5cc85097d
  6. FeedForward model · level 1 · ran · a78cda64d88bd1f2
  7. PreNorm helper · level 1 · ran · 005532cda6b215dd
  8. Residual helper · level 1 · ran · c824fe1a650a3f66
  9. WindowAttention helper · level 1 · ran · 6322ff428c4c429a
  10. Attention model · level 1 · ran · 3bd6d4930cb3f2f8
  11. get_2d_sincos_pos_embed_from_grid helper · level 0 · unverified · 4fd80de79832745d
  12. CyclicShift helper · level 1 · ran · b20098ca471958b8
  13. create_mask helper · level 1 · ran · violated contract · 5b1dded6034b41e3
  14. get_relative_distances helper · level 1 · ran · honoured contract · ea931f8eaca3c48c
  15. get_1d_sincos_pos_embed_from_grid helper · level 1 · ran; the contract check could not run · 4bdd36eab04c3f62

harvested Assembly for Audio-MAE (graph method: Audio-MAE). 7 bricks; individually on a fixture: 1 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 6 unverified; the one sample that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.

one of 4 siblings

  1. MaskedAutoencoderViT entry · level 0 · unverified · 28128850f3b7974d
  2. PatchEmbed_new helper · level 0 · unverified · d28d7b155c97688a
  3. PatchEmbed_org helper · level 1 · ran · aaa7fea64dd2b57c
  4. get_2d_sincos_pos_embed helper · level 0 · unverified · 43ab33710deb02f8
  5. get_2d_sincos_pos_embed_flexible helper · level 0 · unverified · 728825dfce4fad97
  6. get_2d_sincos_pos_embed_from_grid helper · level 0 · unverified · f10004e059714d42
  7. get_1d_sincos_pos_embed_from_grid helper · level 0 · unverified · 12035a2f77d8016c

harvested Assembly for Audio-MAE (graph method: Audio-MAE). 21 bricks; individually on a fixture: 12 ran (of which 4 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 0 violated, 7 with no contract checked; 5 where Syntology's instrument failed) · 9 unverified. The assembly has not been run end-to-end.

one of 4 siblings

  1. audio_mae entry · level 0 · unverified · 8efb3224591920ef
  2. MaskedAutoEncoder model · level 0 · unverified · 680afbcfbca11fb9
  3. SwinTransformer helper · level 0 · unverified · fef623fdfcfbe69b
  4. TransformerEncoder helper · level 0 · unverified · fd67d2fafb66524c
  5. DecoderEmbeddings model · level 1 · ran · 50340b2a447969fb
  6. MAEOutput model · level 0 · unverified · a848037ef9265b59
  7. PatchEmbeddings helper · level 1 · ran · 8163b722029972c0
  8. get_2d_sin_cos_embeddings helper · level 1 · ran; the contract check could not run · 931261dd3b92624d
  9. SwinTransformerBlock helper · level 0 · unverified · 599cc34da67f0356
  10. TransformerOutput helper · level 0 · unverified · f7d645cfcd41319b
  11. Fp32LayerNorm helper · level 1 · ran · 1851c02d87765a1f
  12. TransformerEncoderLayer helper · level 0 · unverified · 7a00feceab29852a
  13. PatchEmbeddingsOutput helper · level 1 · ran · bb30b6d29305dccf
  14. random_masking helper · level 1 · ran; the contract check could not run · 6c5d9dcc5cf2ba42
  15. random_masking_2d helper · level 1 · ran; the contract check could not run · ba9d7aee607505ff
  16. get_1d_sin_cos_embeddings helper · level 1 · ran · our draft was wrong · 3dc442f442bab69d
  17. MLP helper · level 1 · ran · 084f9fe01c1b3cf0
  18. WindowMultiHeadAttention helper · level 0 · unverified · bcffcc5e0273f4c1
  19. MultiHeadSelfAttention helper · level 1 · ran · 354dccecc7d137d8
  20. RandomMaskingOutput helper · level 1 · ran · c2ccb58b26b80780
  21. _random_masking_1d helper · level 1 · ran; the contract check could not run · cad22d0081516d4a

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="2207.06405") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Audio ClassificationDecoderRepresentation LearningSpeaker Identification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Speaker Identification VoxCeleb1 AudioMAE (local) Accuracy 94.8 #4 of 12 Archive leaderboard report
Speaker Identification VoxCeleb1 AudioMAE (local) Top-1 (%) 94.8 #4 of 12 Archive leaderboard report
Speaker Identification VoxCeleb1 AudioMAE (global) Accuracy 94.1 #7 of 12 Archive leaderboard report
Speaker Identification VoxCeleb1 AudioMAE (global) Top-1 (%) 94.1 #7 of 12 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

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutLabel SmoothingLayer NormalizationLinear LayerMAEMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

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