Name: TIGER speech
Author: JusperLee

Overview

TIGER Speech is an audio-to-audio model designed for low-latency, end-to-end speech processing. Unlike traditional pipelines that chain separate ASR and TTS modules, TIGER operates directly on audio signals, significantly reducing the 'robotic' lag and preserving emotional nuance in voice synthesis. For developers, this means a streamlined integration path for building real-time voice assistants or interactive AI agents. Released under the Apache-2.0 license, it offers the flexibility needed for commercial scaling and custom fine-tuning. It is particularly effective for use cases where prosody and natural conversational flow are critical, positioning it as a high-efficiency alternative to fragmented speech stacks.

Highlights

End-to-end audio-to-audio processing for lower latency
Apache-2.0 license allows flexible commercial deployment
Preserves natural prosody better than ASR-TTS pipelines
Optimized for real-time interactive voice applications

Usage

Install

# Install Hugging Face transformers
pip install transformers torch

SDK Usage

# Load model with transformers
from transformers import AutoModel, AutoTokenizer

model = AutoModel.from_pretrained("JusperLee/TIGER-speech")
tokenizer = AutoTokenizer.from_pretrained("JusperLee/TIGER-speech")

Hugging Face Download

We recommend downloading the model via the Hugging Face CLI or Hub SDK.

Guidance：Before downloading, install huggingface_hub with:

Guidance

pip install -U huggingface_hub

CLI Download

Download the full repository

huggingface-cli download JusperLee/TIGER-speech

Download a single file to a local folder (e.g. config.json into ./dir)

huggingface-cli download JusperLee/TIGER-speech config.json --local-dir ./dir

See the official docs for more CLI options

SDK Download

# 模型下载
from huggingface_hub import snapshot_download
model_dir = snapshot_download('JusperLee/TIGER-speech')

Git Download

Make sure git-lfs is installed first

Git Download

git lfs install
git clone https://huggingface.co/JusperLee/TIGER-speech

To skip LFS large-file downloads, use:

Skip LFS

GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/JusperLee/TIGER-speech

Model files are hosted on the Hugging Face Hub — download directly via HF CLI / SDK / Git, not through this site.

PyTorch / Transformers Usage

Install Transformers

pip install -U transformers torch

Load the model and run inference

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained('JusperLee/TIGER-speech')
tokenizer = AutoTokenizer.from_pretrained('JusperLee/TIGER-speech')

Full Documentation

来源: HuggingFace

---
pipeline_tag: audio-to-audio
tags:

audio

license: apache-2.0
language:

en

---

<h3 align="center">TIGER: Time-frequency Interleaved Gain Extraction and Reconstruction for Efficient Speech Separation</h3>

Mohan Xu*, Kai Li*, Guo Chen, Xiaolin Hu 
Tsinghua University, Beijing, China 
*Equal contribution 
<a href="https://arxiv.org/abs/2410.01469">📜 ICLR 2025</a> | <a href="https://cslikai.cn/TIGER/">🎶 Demo</a> | <a href="https://huggingface.co/datasets/JusperLee/EchoSet">🤗 Dataset</a>

> TIGER is a lightweight model for speech separation which effectively extracts key acoustic features through frequency band-split, multi-scale and full-frequency-frame modeling.

💥 News

[2025-01-23] We release the code and pre-trained model of TIGER! 🚀

[2025-01-23] We release the TIGER model and the EchoSet dataset! 🚀

📜 Abstract

In this paper, we propose a speech separation model with significantly reduced parameter size and computational cost: Time-Frequency Interleaved Gain Extraction and Reconstruction Network (TIGER). TIGER leverages prior knowledge to divide frequency bands and applies compression on frequency information. We employ a multi-scale selective attention (MSA) module to extract contextual features, while introducing a full-frequency-frame attention (F^3A) module to capture both temporal and frequency contextual information. Additionally, to more realistically evaluate the performance of speech separation models in complex acoustic environments, we introduce a novel dataset called EchoSet. This dataset includes noise and more realistic reverberation (e.g., considering object occlusions and material properties), with speech from two speakers overlapping at random proportions. Experimental results demonstrated that TIGER significantly outperformed state-of-the-art (SOTA) model TF-GridNet on the EchoSet dataset in both inference speed and separation quality, while reducing the number of parameters by 94.3% and the MACs by 95.3%. These results indicate that by utilizing frequency band-split and interleaved modeling structures, TIGER achieves a substantial reduction in parameters and computational costs while maintaining high performance. Notably, TIGER is the first speech separation model with fewer than 1 million parameters that achieves performance close to the SOTA model.

TIGER

Overall pipeline of the model architecture of TIGER and its modules.

!TIGER Model Architecture

Results

Performance comparisons of TIGER and other existing separation models on *Libri2Mix, LRS2-2Mix, and EchoSet*. Bold indicates optimal performance, and italics indicate suboptimal performance.

!TIGER Model Architecture

Efficiency comparisons of TIGER and other models.

!TIGER Model Architecture

Comparison of performance and efficiency of cinematic sound separation models on DnR. '*' means the result comes from the original paper of DnR.

!TIGER Model Architecture

📦 Installation

bash

git clone https://github.com/JusperLee/TIGER.git
cd TIGER
pip install -r requirements.txt

🚀 Quick Start

Test with Pre-trained Model

bash

# Test using speech
python inference_speech.py --audio_path test/mix.wav
Test using DnR
python inference_dnr.py --audio_path test/test_mixture_466.wav

Train with EchoSet

bash

python audio_train.py --conf_dir configs/tiger.yml

Evaluate with EchoSet

bash

python audio_test.py --conf_dir configs/tiger.yml

📖 Citation

bibtex

@article{xu2024tiger,
  title={TIGER: Time-frequency Interleaved Gain Extraction and Reconstruction for Efficient Speech Separation},
  author={Xu, Mohan and Li, Kai and Chen, Guo and Hu, Xiaolin},
  journal={arXiv preprint arXiv:2410.01469},
  year={2024}
}

📧 Contact

If you have any questions, please feel free to contact us via [email protected].