snakers4/silero-vad
Silero VAD: pre-trained enterprise-grade Voice Activity Detector
Thành tựu nổi bật
Những danh hiệu mới nhất của dự án
#1 Repository Của Tuần
#1 Repository Của Tuần
#1 Repository Của Tuần
🗣️ Silero VAD: Accurate, Lightweight Voice Activity Detection
Pre-trained VAD with stellar accuracy, sub-millisecond latency, and support for PyTorch and ONNX runtimes.
Silero VAD is a pre-trained Voice Activity Detector (VAD) developed by the Silero team. It identifies speech regions in audio, useful for voice interfaces, data cleaning, telephony, and edge/IoT applications. Trained on corpora covering over 6000 languages, it handles diverse noise levels and domains.
✨ Highlights
- Stellar accuracy: Excellent results in speech detection tasks.
- Fast: Processes 30+ ms audio chunks in under 1ms on a single CPU thread.
- Lightweight: JIT model is around two megabytes in size.
- General: Trained on over 6000 languages; works across domains and noise levels.
- Flexible sampling rate: Supports 8000 Hz and 16000 Hz sampling rates.
- Highly portable: Runs on PyTorch and ONNX runtimes across platforms.
- No strings attached: MIT license, no telemetry, keys, or registration.
💡 Why it matters
Silero VAD provides a ready-to-use, accurate model without requiring training. Its small footprint (~2MB) and sub-ms latency make it ideal for edge/IoT and real-time voice applications. ONNX support enables deployment on various architectures beyond x86-64.
#VoiceActivityDetection #PyTorch #ONNX #OpenSource #TiniX
Dự án tương tự
Khám phá các dự án và mô hình AI khác có sự tương đồng.
Using this open-source model in production? Consider switching to pyannoteAI(https://www.pyannote.ai) for better and faster options. 🎹 "Powerset" speaker segmentation This model ingests 10 seconds of mono audio sampled at 16kHz and outputs speaker d...
Using this open-source model in production? Consider switching to pyannoteAI(https://www.pyannote.ai) for better and faster options. 🎹 Speaker segmentation Paper(http://arxiv.org/abs/2104.04045) | Demo(https://huggingface.co/spaces/pyannote/pretrain...
Voice Activity Detector (VAD) : low-latency, high-performance and lightweight
Frame-VAD Multilingual MarbleNet v2.0 Description: Frame-VAD Multilingual MarbleNet v2.0 is a convolutional neural network for voice activity detection (VAD) that serves as the first step for Speech Recognition and Speaker Diarization. It is a frame...
all-MiniLM-L6-v2 This is a sentence-transformers(https://www.SBERT.net) model: It maps sentences & paragraphs to a 384 dimensional dense vector space and can be used for tasks like clustering or semantic search. Usage (Sentence-Transformers) Using t...
🐸💬 - a deep learning toolkit for Text-to-Speech, battle-tested in research and production