audio_gallery.md

Audio Gallery

• Audio Gallery
  • Survey
  • Detection
  • Speech Translation
  • Audio Visual
  • Event Detection
  • Emotion Recognition
  • Audio Separation
  • Diarization
  • Tutorials
  • Toolkits
  • Datasets
  • Products
  • Misc

Survey

Detection

Speech Translation

Audio Visual

Event Detection

• CLIP-VAD: Exploiting Vision-Language Models for Voice Activity Detection, arXiv, 2410.14509, arxiv, pdf, cication: -1

  Andrea Appiani, Cigdem Beyan

Emotion Recognition

Audio Separation

• Vocal Track Separation with Encoder-Decoder Architecture

  · (keras-io - keras-team)

• ClearerVoice-Studio - modelscope

• svoice - facebookresearch

  Speaker Voice Separation using Neural Nets · (arxiv)

• 掌握音频分离黑科技，这几款工具让你的音频瞬间焕然一新！

  · (vocal-separate - jianchang512) · (bilibili) · (ultimatevocalremovergui - Anjok07)

Diarization

• 3D-Speaker - modelscope
• DiariZen - BUTSpeechFIT

Tutorials

Toolkits

• Speaker Verification with ECAPA-TDNN embeddings on Voxceleb 🤗
• wavesurfer - pengzhendong

Datasets

• 🌟 AudioSetCaps: An Enriched Audio-Caption Dataset using Automated Generation Pipeline with Large Audio and Language Models, arXiv, 2411.18953, arxiv, pdf, cication: -1

  Jisheng Bai, Haohe Liu, Mou Wang, ..., Woon-Seng Gan, Jianfeng Chen · (AudioSetCaps - JishengBai)

• LAION-DISCO-12M

  · (huggingface)

• The Expresso dataset is a high-quality (48kHz) expressive speech dataset that includes both expressively rendered read speech 🤗

  · (speechbot.github)

• SonicSim: A customizable simulation platform for speech processing in moving sound source scenarios, arXiv, 2410.01481, arxiv, pdf, cication: -1

  Kai Li, Wendi Sang, Chang Zeng, ..., Guo Chen, Xiaolin Hu · (cslikai) · (SonicSim - JusperLee) · (mp.weixin.qq)

• NCSSD dataset and collecting pipeline to handle TV shows. 🤗

• 电台直播

• EmotionCaps: Enhancing Audio Captioning Through Emotion-Augmented Data Generation, arXiv, 2410.12028, arxiv, pdf, cication: -1

  Mithun Manivannan, Vignesh Nethrapalli, Mark Cartwright

Products

Misc

• versa - shinjiwlab

• SUPERB is a collection of benchmarking resources to evaluate the capability of a universal shared representation for speech processing.

• Vision Language Models Are Few-Shot Audio Spectrogram Classifiers, arXiv, 2411.12058, arxiv, pdf, cication: -1

  Satvik Dixit, Laurie M. Heller, Chris Donahue

• llama-recipes - meta-llama

  An Open Source version of NotebookLM