Project is maintened and described in my blog. Check it!
VoiceHAC (Voice Home Automation Controller) is a comprehensive solution for controlling smart home devices through natural language processing. The project uses custom-trained NER models to understand voice commands, enabling users to control various aspects of their smart home environment.
VoiceHAC-NER (Voice Home Automation Controller - Named Entity Recognition) is a project that uses natural language processing techniques to parse and understand voice commands for controlling smart home devices. The project leverages a custom-trained named entity recognition model using the SpaCy library to identify relevant entities in the given text.
- Custom-trained NER model to recognize home automation-related entities
- Named entity recognition for actions, attributes, things, and locations
- Extract numerical values, including fractions and percentages
- Install the required dependencies with
pip:
```sh pip install -r requirements.txt ```
- Download and install the pre-trained SpaCy model (if not already installed):
```sh python -m spacy download en_core_web_sm ```
- Configure the paths in
config.pyto point to your trained model and test sentences file.
Import the VoiceHAC_NER class in your Python script or application, and use it to process text commands:
```python from voicehac_ner import VoiceHAC_NER
ner = VoiceHAC_NER(model_path) entities = ner.get_named_entities(text) numbers = ner.extract_numerical_values(text) ```
Replace model_path with the path to your custom-trained model, and text with the input command.
This project is licensed under the MIT License.