Papers › Conv-TasNet: Surpassing Ideal Time-Frequency Magnitude Masking for Speech Separation

Conv-TasNet: Surpassing Ideal Time-Frequency Magnitude Masking for Speech Separation

20 Sep 2018arXiv:1809.07454archive 2025-07-28

Yi Luo, Nima Mesgarani

Single-channel, speaker-independent speech separation methods have recently seen great progress. However, the accuracy, latency, and computational cost of such methods remain insufficient. The majority of the previous methods have formulated the separation problem through the time-frequency representation of the mixed signal, which has several drawbacks, including the decoupling of the phase and magnitude of the signal, the suboptimality of time-frequency representation for speech separation, and the long latency in calculating the spectrograms. To address these shortcomings, we propose a fully-convolutional time-domain audio separation network (Conv-TasNet), a deep learning framework for end-to-end time-domain speech separation. Conv-TasNet uses a linear encoder to generate a representation of the speech waveform optimized for separating individual speakers. Speaker separation is achieved by applying a set of weighting functions (masks) to the encoder output. The modified encoder representations are then inverted back to the waveforms using a linear decoder. The masks are found using a temporal convolutional network (TCN) consisting of stacked 1-D dilated convolutional blocks, which allows the network to model the long-term dependencies of the speech signal while maintaining a small model size. The proposed Conv-TasNet system significantly outperforms previous time-frequency masking methods in separating two- and three-speaker mixtures. Additionally, Conv-TasNet surpasses several ideal time-frequency magnitude masks in two-speaker speech separation as evaluated by both objective distortion measures and subjective quality assessment by human listeners. Finally, Conv-TasNet has a significantly smaller model size and a shorter minimum latency, making it a suitable solution for both offline and real-time speech separation applications.

PaperPDFCodeCode 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

The archive's official repository (naplab/Conv-TasNet) was not harvested; the 36 samples below come from 7 community repositories.

Syntology 36 code samples harvested from 7 repositories linked to this paper: 28 ran (of which 0 constructed an object rather than computing a result; 22 with no instrument failure: 3 honoured, 2 violated, 17 with no contract checked; 6 where Syntology's instrument failed) · 8 unverified. Where Syntology's instrument failed: 5 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: community (archive-listed): 36 samples from 7 repositories, 28 ran (22 with no instrument failure; 6 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.

No sample is linked to this paper from 5 of its repositories that Syntology harvested for other papers:

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.

17 repositories listed; official and paper-mentioned ones first.

naplab/Conv-TasNet officialmentioned in paperpytorchSyntology: not harvested report
IliasDG/Demucs_Project mentioned on GitHubpytorchMITSyntology: 13 samples, 11 ran (4 instrument-failed) · 2 unverified report
adamwhitakerwilson/speaker_separation mentioned on GitHubpytorchSyntology: not harvested report
alibabasglab/frcrn mentioned on GitHubpytorchSyntology: not harvested report
facebookresearch/demucs mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Abusive music and song transformation using GenAI and LLMs (arXiv:2601.15348; that paper's own run record: 7 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 3 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
facebookresearch/svoice mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Voice Separation with an Unknown Number of Multiple Speakers (arXiv:2003.01531; that paper's own run record: 12 ran (of which 5 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 1 violated, 8 with no contract checked; 2 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
funcwj/conv-tas-net mentioned on GitHubpytorchMITSyntology: not harvested report
funcwj/conv-tasnet mentioned on GitHubpytorchMITSyntology: 10 samples, 8 ran (0 instrument-failed) · 2 unverified report
kaituoxu/Conv-TasNet mentioned on GitHubpytorchSyntology: 3 samples, 3 ran (1 instrument-failed) · 0 unverified report
mitmedialab/unet4birds mentioned on GitHubpytorchSyntology: not harvested report
mpariente/asteroid mentioned on GitHubpytorchMITSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to DCCRN: Deep Complex Convolution Recurrent Network for Phase-Aware Speech Enhancement (arXiv:2008.00264; that paper's own run record: 14 ran (of which 0 constructed an object rather than computing a result; 14 with no instrument failure: 4 honoured, 0 violated, 10 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
muhammad-ahmed-ghani/svoice_demo mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Voice Separation with an Unknown Number of Multiple Speakers (arXiv:2003.01531; that paper's own run record: 12 ran (of which 5 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 1 violated, 8 with no contract checked; 2 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
pfnet-research/meta-tasnet mentioned on GitHubpytorchMITSyntology: 5 samples, 2 ran (0 instrument-failed) · 3 unverified report
robieta/demucs mentioned on GitHubpytorchMITSyntology: 1 sample, 0 ran · 1 unverified report
saurjya/asteroid mentioned on GitHubpytorchMITSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Dual-Path Transformer Network: Direct Context-Aware Modeling for End-to-End Monaural... (arXiv:2007.13975; that paper's own run record: 7 ran (of which 0 constructed an object rather than computing a result; 7 with no instrument failure: 1 honoured, 1 violated, 5 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
JusperLee/Conv-TasNet pytorchSyntology: 2 samples, 2 ran (1 instrument-failed) · 0 unverified report
xiaolianggeitazgh/Conv-Tasnet mindsporeSyntology: 2 samples, 2 ran (0 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.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

36 samples harvested: 28 ran (of which 0 constructed an object rather than computing a result; 22 with no instrument failure: 3 honoured, 2 violated, 17 with no contract checked; 6 where Syntology's instrument failed) · 8 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

3ran · honoured contract
2ran · violated contract
5ran · our draft was wrong
1ran; the contract check could not run
17ran
8unverified

Licence: 16 of the 36 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 7 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”. “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.

IliasDG/Demucs_Project community (archive-listed) 13 samples, 11 ran (4 instrument-failed) · 2 unverified
average_metric demucs/utils.py ran MIT recorded; this copy not marked cleared · pointer only · a68dbf50b2629a55 · report fixture: signature metric:float/0/float, count:float/0/float · returned float · level recorded 2026-09-25
center_trim in 5 papers demucs/utils.py ran fingerprinted MIT recorded; this copy not marked cleared · pointer only · f89970356c8e9506 · report fixture: signature tensor:Tensor/3/float, reference:Tensor/3/float · returned Tensor · level recorded 2026-09-25
chose_norm in 5 papers demucs/tasnet.py ran · our draft was wrong MIT recorded; this copy not marked cleared · pointer only · c20ada1396105c22 · report fixture: signature norm_type:str/0/str, channel_size:int/0/int · returned GlobalLayerNorm · level recorded 2026-09-25
downsample in 59 papers demucs/model.py ran; the contract check could not run (RAISES) fingerprinted MIT recorded; this copy not marked cleared · pointer only · 07fa2ca4b8bf0830 · report fixture: signature x:Tensor/3/float, stride:int/0/int · returned Tensor · level recorded 2026-09-25
fname dora.py ran MIT recorded; this copy not marked cleared · pointer only · 9f54a0c5172ff77c · report fixture: signature name:str/0/str, kind:str/0/str · returned PosixPath · level recorded 2026-09-26
free_port demucs/utils.py ran MIT recorded; this copy not marked cleared · pointer only · f747e6cf6fc620a6 · report fixture: signature host:str/0/str, low:int/0/int, high:int/0/int · returned int · level recorded 2026-09-25
get_done dora.py ran MIT recorded; this copy not marked cleared · pointer only · 7bd6c5376583922a · report fixture: signature name:str/0/str · returned bool · level recorded 2026-09-26
get_name in 20 papers demucs/parser.py ran MIT recorded; this copy not marked cleared · pointer only · dccae3e25e1c55a2 · report fixture: signature parser:MockParser/0/?, args:MockArgs/0/? · returned str · level recorded 2026-09-25
overlap_and_add in 8 papers demucs/tasnet.py ran · our draft was wrong fingerprinted MIT recorded; this copy not marked cleared · pointer only · 6fc6cdb0729e7905 · report fixture: signature signal:Tensor/3/float, frame_step:int/0/int · returned Tensor · level recorded 2026-09-25
quantize in 70 papers
census: 8 implementations from 11 papers, 6 distinct outputs across 3 buckets, one shared input per bucket
quantize.py ran fingerprinted MIT recorded; this copy not marked cleared · pointer only · 95678855edc371ba · report fixture: signature p:Tensor/2/float, level:int/0/int · returned Tensor · level recorded 2026-09-26
upsample in 87 papers
census: 11 implementations from 16 papers, 9 distinct outputs across 3 buckets, one shared input per bucket
demucs/model.py ran · our draft was wrong fingerprinted MIT recorded; this copy not marked cleared · pointer only · 5cf51541da88d14a · report fixture: signature x:Tensor/3/float, stride:int/0/int · returned Tensor · level recorded 2026-09-25
build_metadata demucs/compressed.py unverified MIT recorded; this copy not marked cleared · pointer only · ffd24b3bb5a4d744 · report
get_sid dora.py unverified MIT recorded; this copy not marked cleared · pointer only · 4678393521f60867 · report
funcwj/conv-tasnet community (archive-listed) 10 samples, 8 ran (0 instrument-failed) · 2 unverified
build_norm in 5 papers nnet/conv_tas_net.py ran MIT (permissive) · cc51e06c5fd96123 · report fixture: signature norm:str/0/str, dim:int/0/int · returned ChannelWiseLayerNorm · level recorded 2026-09-25
get_logger in 383 papers nnet/libs/utils.py ran MIT (permissive) · d0a0123e9bc8a4c4 · report fixture: signature name:str/0/str, format_str:str/0/str, date_format:str/0/str, file:bool/0/bool · returned Logger · level recorded 2026-09-25
load_json in 344 papers nnet/libs/utils.py ran MIT (permissive) · 1a0aeff1eb7546d6 · report fixture: signature fdir:str/0/str, name:str/0/str · returned dict · level recorded 2026-09-25
load_obj in 34 papers nnet/libs/trainer.py ran MIT (permissive) · 9b384ef1d2d3ea3f · report fixture: signature obj:dict/0/?, device:str/0/str · returned dict · level recorded 2026-09-25
param nnet/conv_tas_net.py ran MIT (permissive) · 3a4ca7c24b1659fc · report fixture: signature nnet:callable/0/?, Mb:bool/0/bool · returned float · level recorded 2026-09-25
parse_scripts nnet/libs/audio.py ran MIT (permissive) · 77ec6163bd018420 · report fixture: signature scp_path:str/0/str, value_processor:callable/0/?, num_tokens:int/0/int · returned dict · level recorded 2026-09-25
permute_si_snr nnet/libs/metric.py ran MIT (permissive) · 457f6fe40f154950 · report fixture: signature xlist:list/1/ndarray, slist:list/1/ndarray · returned float64 · level recorded 2026-09-25
si_snr nnet/libs/metric.py ran MIT (permissive) · 23b2bdd3e0fe81c3 · report fixture: signature x:ndarray/1/float, s:ndarray/1/float, remove_dc:bool/0/bool · returned float32 · level recorded 2026-09-25
make_dataloader in 18 papers nnet/libs/dataset.py unverified MIT (permissive) · 4e23d73a9b135dab · report
read_wav nnet/libs/audio.py unverified MIT (permissive) · 455d88ef82f00d25 · report
kaituoxu/Conv-TasNet community (archive-listed) 3 samples, 3 ran (1 instrument-failed) · 0 unverified
cal_SISNR src/evaluate.py ran · violated contract fingerprinted MIT (permissive) · faac9c0b7e1a08c5 · report fixture: signature ref_sig:ndarray/1/float, out_sig:ndarray/1/float, eps:float/0/float · returned float64 · level recorded 2026-09-22
cal_SISNRi in 5 papers src/evaluate.py ran · honoured contract fingerprinted MIT (permissive) · b19791aa06bb7b9b · report fixture: signature src_ref:ndarray/2/float, src_est:ndarray/2/float, mix:ndarray/1/float · returned float64 · level recorded 2026-09-22
chose_norm in 5 papers src/conv_tasnet.py ran · our draft was wrong MIT (permissive) · 22f921651bb6e769 · report fixture: signature norm_type:str/0/str, channel_size:int/0/int · returned GlobalLayerNorm · level recorded 2026-09-22
pfnet-research/meta-tasnet community (archive-listed) 5 samples, 2 ran (0 instrument-failed) · 3 unverified
sdr_objective utility/loss.py ran MIT (permissive) · 175694073fa2797e · report fixture: signature estimation:Tensor/4/float, origin:Tensor/4/float, mask:Tensor/3/bool · returned Tensor · level recorded 2026-09-26
similarity_loss in 5 papers utility/loss.py ran MIT (permissive) · 865adc65cb86cc62 · report fixture: signature latents:Tensor/4/float, mask:Tensor/2/bool · returned Tensor · level recorded 2026-09-26
Conv1dWrapper model/conv1d.py unverified MIT (permissive) · b80414a2c961fe63 · report
GroupNormWrapper model/group_norm.py unverified MIT (permissive) · 60382cdca01c9e16 · report
dissimilarity_loss utility/loss.py unverified MIT (permissive) · 6d72bd968efb3b87 · report
robieta/demucs community (archive-listed) 1 sample, 0 ran · 1 unverified
build_metadata demucs/compressed.py unverified MIT recorded; this copy not marked cleared · pointer only · b482cd70c4311bcc · report
JusperLee/Conv-TasNet community (archive-listed) 2 samples, 2 ran (1 instrument-failed) · 0 unverified
check_parameters Conv_TasNet_Pytorch/Conv_TasNet.py ran · honoured contract no licence file found · pointer only · bbd39ee3588090e3 · report fixture: signature net:callable/0/? · returned float · level recorded 2026-09-22
select_norm in 6 papers Conv_TasNet_Pytorch/Conv_TasNet.py ran · our draft was wrong no licence file found · pointer only · fd365b69d72fe0c9 · report fixture: signature norm:str/0/str, dim:int/0/int · returned GlobalLayerNorm · level recorded 2026-09-22
xiaolianggeitazgh/Conv-Tasnet community (archive-listed) 2 samples, 2 ran (0 instrument-failed) · 0 unverified
cal_SISNR evaluate.py ran · violated contract fingerprinted Apache-2.0 (permissive) · 891e417fc38edeb0 · report fixture: signature ref_sig:ndarray/1/float, out_sig:ndarray/1/float, eps:float/0/float · returned float64 · level recorded 2026-09-22
cal_SISNRi in 5 papers evaluate.py ran · honoured contract fingerprinted Apache-2.0 (permissive) · ca6c3339d98f2a2c · report fixture: signature src_ref:ndarray/2/float, src_est:ndarray/2/float, mix:ndarray/1/float · returned float64 · level recorded 2026-09-22

Tasks

Multi-task Audio Source SeperationMusic Source SeparationSpeaker SeparationSpeech EnhancementSpeech Separation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Music Source Separation MUSDB18 Conv-TasNet (extra) SDR (avg) 6.32 #14 of 27 Archive leaderboard report
Music Source Separation MUSDB18 Conv-TasNet (extra) SDR (bass) 7.00 #14 of 27 Archive leaderboard report
Music Source Separation MUSDB18 Conv-TasNet (extra) SDR (drums) 7.11 #14 of 27 Archive leaderboard report
Music Source Separation MUSDB18 Conv-TasNet (extra) SDR (vocals) 6.74 #14 of 27 Archive leaderboard report
Music Source Separation MUSDB18 Conv-TasNet SDR (avg) 5.73 #22 of 27 Archive leaderboard report
Music Source Separation MUSDB18 Conv-TasNet SDR (bass) 5.66 #22 of 27 Archive leaderboard report
Music Source Separation MUSDB18 Conv-TasNet SDR (drums) 6.08 #22 of 27 Archive leaderboard report
Music Source Separation MUSDB18 Conv-TasNet SDR (other) 4.37 #22 of 27 Archive leaderboard report
Music Source Separation MUSDB18 Conv-TasNet SDR (vocals) 6.81 #22 of 27 Archive leaderboard report
Speech Enhancement EARS-WHAM Conv-TasNet DNSMOS 3.47 #5 of 6 Archive leaderboard report
Speech Enhancement EARS-WHAM Conv-TasNet ESTOI 0.70 #5 of 6 Archive leaderboard report
Speech Enhancement EARS-WHAM Conv-TasNet PESQ-WB 2.31 #5 of 6 Archive leaderboard report
Speech Enhancement EARS-WHAM Conv-TasNet POLQA 2.73 #5 of 6 Archive leaderboard report
Speech Enhancement EARS-WHAM Conv-TasNet SI-SDR 16.93 #5 of 6 Archive leaderboard report
Speech Enhancement EARS-WHAM Conv-TasNet SIGMOS 2.69 #5 of 6 Archive leaderboard report
Speech Separation WSJ0-2mix Conv-TasNet Number of parameters (M) 5.1 #36 of 40 Archive leaderboard report
Speech Separation WSJ0-2mix Conv-TasNet SDRi 15.6 #36 of 40 Archive leaderboard report
Speech Separation WSJ0-2mix Conv-TasNet SI-SDRi 15.3 #36 of 40 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

ConvTasNet

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