🧠 soundKIT Voice Activity Detection (VAD) Tutorial¶

This tutorial walks you through using the soundKIT CLI for VAD tasks including data preparation, training, evaluation, TFLite export, and embedded deployment.

In [ ]:

# ⚙️ Setup
# Set working directory to project root (one level up)
import os
os.chdir("..")
print("✅ Working directory set to project root:", os.getcwd())

# ⚙️ Setup # Set working directory to project root (one level up) import os os.chdir("..") print("✅ Working directory set to project root:", os.getcwd())

In [ ]:

# Clone the soundKIT repo
!git clone https://github.com/AmbiqAI/soundkit.git
%cd soundkit

# Clone the soundKIT repo !git clone https://github.com/AmbiqAI/soundkit.git %cd soundkit

In [ ]:

# Install required dependencies
!pip install -e .

# Install required dependencies !pip install -e .

🔧 Configuration¶

In [ ]:

# Show editable VAD training config
config_path = "configs/vad.yaml"
with open(config_path, "r") as f:
    print(f.read())

# Show editable VAD training config config_path = "configs/vad.yaml" with open(config_path, "r") as f: print(f.read())

📁 Step 1: Prepare Data¶

In [ ]:

!soundkit -t vad -m data -c configs/vad.yaml

!soundkit -t vad -m data -c configs/vad.yaml

🎯 Step 2: Train Model¶

In [ ]:

!soundkit -t vad -m train -c configs/vad.yaml

!soundkit -t vad -m train -c configs/vad.yaml

To monitor training with TensorBoard:

In [ ]:

!soundkit -t vad -m train --tensorboard -c configs/vad.yaml

!soundkit -t vad -m train --tensorboard -c configs/vad.yaml

Then run this in another terminal:

In [ ]:

# tensorboard --logdir=soundkit/tasks/vad/tensorboard/your_run

# tensorboard --logdir=soundkit/tasks/vad/tensorboard/your_run

📊 Step 3: Evaluate Model¶

In [ ]:

!soundkit -t vad -m evaluate -c configs/vad.yaml

!soundkit -t vad -m evaluate -c configs/vad.yaml

📤 Step 4: Export Model to TFLite¶

In [ ]:

!soundkit -t vad -m export -c configs/vad.yaml

!soundkit -t vad -m export -c configs/vad.yaml

🚀 Step 5: Deploy to Embedded EVB (e.g., Apollo)¶

In [ ]:

!soundkit -t vad -m demo -c configs/vad.yaml # For EVB demo, ensure you have a microphone connected
!soundkit -t vad -m demo -c configs/vad.yaml demo.platform=pc # For PC demo, ensure you have a microphone connected

!soundkit -t vad -m demo -c configs/vad.yaml # For EVB demo, ensure you have a microphone connected !soundkit -t vad -m demo -c configs/vad.yaml demo.platform=pc # For PC demo, ensure you have a microphone connected

✅ Notes¶

• Edit vad.yaml to adjust model hyperparameters, features, and output paths
• Log files and detection visualizations are saved under the configured test and training folders