---
license: cc-by-nc-4.0
library_name: libreyolo
tags:
  - object-detection
  - zero-shot-object-detection
  - open-vocabulary-detection
  - libreyolo
---

# LibreOVDEIMs

OV-DEIM-S real-time open-vocabulary detector (11M params, ViT-tiny (DEIMv2 distilled)
backbone) repackaged for LibreYOLO's `LibreOpenVocab` tier, bundled with the
MobileCLIP-B(LT) text tower for online prompt encoding.

## Source

Detector converted from [wleilei/OV-DEIM](https://github.com/wleilei/OV-DEIM)
at commit `dfbf394672407b7f837ec08e7d68e8127548b254` (checkpoint
`ovdeim_s.pth`, trained on Objects365v1 + GoldG). Text encoder taken from
[apple/MobileCLIP-B-LT-OpenCLIP](https://huggingface.co/apple/MobileCLIP-B-LT-OpenCLIP).

Paper: OV-DEIM: Real-time DETR-Style Open-Vocabulary Object Detection with
GridSynthetic Augmentation (arXiv 2603.07022). Upstream reports 161 FPS on a
T4 with TensorRT at 640 input.

## Modifications

State-dict key remapping and merge only (`module.` prefix stripped, the
training-only denoising embedding dropped, the MobileCLIP text tower stored
under `text_encoder.*`). Learned parameters are unchanged. See
`weights/convert_ovdeim_weights.py` in the
[LibreYOLO source repository](https://github.com/LibreYOLO/libreyolo).

LibreYOLO's port reproduces the upstream paper's zero-shot COCO val2017
result through this artifact (45.9 AP for the L size).

## Usage

```python
from libreyolo import LibreOpenVocab

model = LibreOpenVocab("ov-deim-s")
model.set_classes(["person", "a red backpack", "forklift"])
results = model.predict("image.jpg")
```

## License

Layered, most restrictive governs the combined artifact:

- Detector weights: [CC BY-NC 4.0](./LICENSE) (non-commercial, attribution;
  redistribution and format conversion permitted by upstream's MODEL_LICENSE).
- OV-DEIM code lineage: [Apache-2.0](./LICENSE_CODE).
- MobileCLIP-B(LT) text tower: [Apple Machine Learning Research Model
  license](./LICENSE_MOBILECLIP_WEIGHTS), research use only.
- DINOv3-derived backbone: [DINOv3 License](./LICENSE_DINOV3).

See [`NOTICE`](./NOTICE) for attribution details.