Wav2vec2 Base Ft Keyword Spotting
W
Wav2vec2 Base Ft Keyword Spotting
Developed by anton-l
A speech keyword recognition model fine-tuned on the SUPERB dataset based on facebook/wav2vec2-base, achieving an accuracy of 98.26%
Downloads 70
Release Time : 3/2/2022
Model Overview
This model is a speech classification model for keyword recognition, fine-tuned based on the wav2vec2 architecture, excelling at detecting specific keywords from speech
Model Features
High Accuracy
Achieves an accuracy of 98.26% on the evaluation set, demonstrating excellent performance
Based on wav2vec2 Architecture
Utilizes the proven wav2vec2-base architecture as the foundational model
Efficient Fine-tuning
Targeted fine-tuning on the SUPERB dataset optimizes keyword recognition capabilities
Model Capabilities
Speech Classification
Keyword Recognition
Audio Feature Extraction
Use Cases
Smart Home
Voice Wake-up
Used for voice wake-up functionality in smart devices
High accuracy in recognizing wake words
Voice Assistants
Command Word Recognition
Identifies key command words in user voice instructions
Featured Recommended AI Models
Š 2025AIbase