Instructions to use LibreYOLO/LibreFOMOl-point with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- LibreYOLO
How to use LibreYOLO/LibreFOMOl-point with LibreYOLO:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
LibreFOMOl-point
LibreFOMO size-l point-localizer (224Γ224 input, 28Γ28 detection grid), packaged for LibreYOLO.
Architecture-only weights. These weights are randomly initialised (nc=1, class "object"). They produce no meaningful predictions out of the box. Fine-tune on your own labelled data before deploying. See the LibreYOLO training guide for instructions.
Architecture
LibreFOMO is a MobileNetV2-derived point-localizer designed for counting and locating objects at very low resolutions. It outputs a spatial grid of per-class confidence scores rather than bounding boxes, making it well-suited for fixed-camera occupancy monitoring and object counting on edge devices.
| Size | Input | Grid | Parameters |
|---|---|---|---|
| l | 224Γ224 | 28Γ28 | ~297 KB |
Usage
from libreyolo import LibreFOMO
# Load architecture weights (random init β fine-tune before use)
model = LibreFOMO("LibreFOMOl-point.pt")
# Fine-tune on your data
model.train(data="your_data.yaml", allow_experimental=True)
License
MIT License. See the LICENSE file in this repository.
- Downloads last month
- 48