Papers › Investigating Training Objectives for Generative Speech Enhancement

Investigating Training Objectives for Generative Speech Enhancement

16 Sep 2024arXiv:2409.10753archive 2025-07-28

Julius Richter, Danilo de Oliveira, Timo Gerkmann

Generative speech enhancement has recently shown promising advancements in improving speech quality in noisy environments. Multiple diffusion-based frameworks exist, each employing distinct training objectives and learning techniques. This paper aims to explain the differences between these frameworks by focusing our investigation on score-based generative models and the Schrödinger bridge. We conduct a series of comprehensive experiments to compare their performance and highlight differing training behaviors. Furthermore, we propose a novel perceptual loss function tailored for the Schrödinger bridge framework, demonstrating enhanced performance and improved perceptual quality of the enhanced speech signals. All experimental code and pre-trained models are publicly available to facilitate further research and development in this domain.

PaperPDFCode

In Syntology View this paper on Syntology, its page in Syntology's graph. That page 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

sp-uhh/sgmse officialmentioned in papermentioned on GitHubpytorchMITSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to EARS: An Anechoic Fullband Speech Dataset Benchmarked for Speech Enhancement and... (arXiv:2406.06185; that paper's own run record: 8 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 1 honoured, 0 violated, 3 with no contract checked; 4 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. 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

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Speech Enhancement

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Speech Enhancement EARS-WHAM Schrödinger Bridge (PESQ loss) DNSMOS 3.72 #1 of 6 Archive leaderboard report
Speech Enhancement EARS-WHAM Schrödinger Bridge (PESQ loss) ESTOI 0.73 #1 of 6 Archive leaderboard report
Speech Enhancement EARS-WHAM Schrödinger Bridge (PESQ loss) PESQ-WB 3.09 #1 of 6 Archive leaderboard report
Speech Enhancement EARS-WHAM Schrödinger Bridge (PESQ loss) POLQA 3.71 #1 of 6 Archive leaderboard report
Speech Enhancement EARS-WHAM Schrödinger Bridge (PESQ loss) SI-SDR 16.29 #1 of 6 Archive leaderboard report
Speech Enhancement EARS-WHAM Schrödinger Bridge (PESQ loss) SIGMOS 3.18 #1 of 6 Archive leaderboard report
Speech Enhancement VoiceBank + DEMAND Schrödinger bridge (PESQ loss) PESQ (wb) 3.70 #4 of 42 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