--- license: apache-2.0 language: - en pipeline_tag: audio-to-audio --- # Regularized Schrödinger Bridge (RSB) via Distortion-Perception Perturbation for High-Fidelity Speech Enhancement [![License](https://img.shields.io/badge/license-Apache%202.0-blue.svg)](LICENSE) Regularized Schrödinger Bridge (RSB) is a generative speech enhancement approach that reconciles fidelity and realism while mitigating exposure bias. RSB regularizes training with a Distortion-Perception Perturbation that constructs time-varying targets by interpolating between clean speech and posterior-mean estimates, and trains the network on perturbed intermediate states to correct toward the ground truth progressively. Consequently, such perturbation simulates inference-time prediction errors, mitigating the training–inference mismatch and thereby reducing exposure bias. Furthermore, it also injects posterior-mean estimates as fidelity-preserving guidance, facilitating reconstruction fidelity. - Official PyTorch implementation of the paper: [Regularized Schrödinger Bridge via Distortion-Perception Perturbation for High-Fidelity Speech Enhancement]() - **Links**: Paper | [Audio Demo](https://yorch233.github.io/RSB/) | [Online Demo](https://huggingface.co/spaces/Yorch233/RSB) | [Github](https://github.com/Yorch233/RSB) | [Huggingface](https://huggingface.co/Yorch233/RSB) #### Pretrained Model Download We have publicly released a checkpoint of MISB's generative model, which is based the `ncsnpp_base` architecture and was trained on the `Voicebank+Demand` dataset. There are two ways to download: + Download via `CLI` ```Bash python -m cli.download_pretrained_model ``` + Download via `Google Drive`. Download the folder from [Google Drive](https://drive.google.com/drive/folders/1b5wI-DIvq3yyLH_PKJPeb3d8VwxfwYIR?usp=sharing) and place it in the `pretrained_models/` directory. ## License This project is licensed under the [Apache License 2.0](LICENSE).