Model reference · open weights
DriftSE is an open-weight audio or speech model from LIANGXU123, listed in the AxForge catalogue. AxForge can bring it up on EU-owned hardware for you on request — with the licence handled where one is required.
About
DriftSE — Speech Enhancement Based on Drifting Models [](https://arxiv.org/abs/2604.24199) [](https://github.com/liangxu123/driftse) [](https://opensource.org/licenses/MIT) []() Speech Enhancement Based on Drifting Models (Interspeech 2026, Oral Presentation) Liang Xu, Diego Caviedes-Nozal, W. Bastiaan Kleijn, Longfei Felix Yan, Rasmus Kongsgaard Olsson 🔗 Project Website | 📄 arXiv Paper | 💻 GitHub Code| 🤗 Hugging Face Space ✨ Key Highlights - Novel Generative Paradigm — Formulates speech enhancement as a distributional equilibrium problem, eliminating the need for iterative denoising or trajectory-based sampling. - Native One-Step Inference — Achieves single-step (1 NFE) enhancement by evolving the pushforward distribution of a mapping function to directly match the clean speech distribution via a Drifting Field. - Semantic Latent Drifting — Operates in a hierarchical self-supervised speech latent space (HuBERT, WavLM, DistilHuBERT), providing rich and stable training signals that capture both acoustic and phonetic structure. - Unpaired Learning — Natively supports training on fully unpaired noisy/clean speech data, enabling cross-dataset and cross-gender generalization without paired supervision. - State-of-the-Art Generalization — Achieves state-of-the-art WV-MOS and SCOREQ on the DNS Challenge 2020 blind test set, outperforming multi-step diffusion and consistency-based baselines. 📦 Repository Contents This Hugging Face repository hosts pre-trained checkpoints and enhanced audio outputs for DriftSE. 📊 Performance Benchmark VoiceBank-DEMAND (VB-DMD) — In-Domain Evaluation † Jointly trained with auxiliary PESQ, SI-SDR, and CCMSE losses. DNS Challenge 2020 Blind Test Set — Real-World Generalization 🚀 Quick Start 1. Install Dependencies 2. Download Checkpoint Or via CLI: 3. Run Enhancement The evaluation pipeline runs two phases: 1. Enhancement — generates enhanced audio via enhancement.py 2. Objective Metrics — computes PESQ, ESTOI, SI-SDR via calcmetrics.py 🏗️ Model Architecture Two Formulations - Conditional Generator (withz/) — Stochastic mapping fθ(ε, y) from Gaussian noise conditioned on noisy speech, optimized for perceptual q
Summarised from the published model card. Read the full card on the HuggingFace links below.
Specifications
| Maker | LIANGXU123 |
|---|---|
| Type | Audio & music |
| Variants | 1 |
| Runs with | pytorch |
| Released | 2026-08-13 |
| Popularity | 1k downloads / month |
| Likes | 1 |
| Licence | Open weights |
How it works
Variants
Open weights ship in several sizes and precisions. One page, all the variants — pick the one that fits your GPU. VRAM figures are estimates from model size.
| Variant | Params | Precision | VRAM | Fits 16 GB | Weights |
|---|---|---|---|---|---|
| DriftSE | — | BF16 | — | — | Weights ↗ |
Using it via the API
Once AxForge deploys driftse for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (driftse below is illustrative; you get the exact model name on deployment.)
$ curl -sS https://api.axforge.ai/v1/audio/transcriptions \ -H "Authorization: Bearer $AXFORGE_API_KEY" \ -F model="driftse" -F file=@audio.mp3
Licence
Open weights under mit — commercial use is permitted. Deploy it on AxForge EU hardware on request. Read the licence ↗