README.md

---
license: mit
library_name: libreyolo
pipeline_tag: depth-estimation
tags:
  - depth-estimation
  - monocular-depth
  - zipdepth
---

# LibreZipDepthb-depth

ZipDepth base: a ~6M-parameter reparameterizable CNN (RepVGG encoder, SPPF +
cross-scale neck, FPN decoder with convex upsampling) for zero-shot relative
monocular depth estimation, repackaged for LibreYOLO. This is the standard
GPU/CPU checkpoint; the sibling repo
[LibreZipDepthbnpu-depth](https://huggingface.co/LibreYOLO/LibreZipDepthbnpu-depth)
carries the separately trained unfold-free decoder for NPU/edge compilers.

## Usage

```python
from libreyolo import LibreYOLO

model = LibreYOLO("LibreZipDepthb-depth.pt")   # auto-downloads from this repo
results = model.predict("image.jpg", save=True)
depth = results[0].depth_map.data              # (H, W) relative inverse depth
```

Outputs follow LibreYOLO's depth task contract: a dense float map on the
original image canvas, higher values mean closer to the camera, no metric unit.

## Source

Derived from [fabiotosi92/ZipDepth](https://github.com/fabiotosi92/ZipDepth)
(`checkpoints/zipdepth_base.pth`) at commit
`6b96f4d205f8a2e5377e81c1b74cc99a47f6693a` (ECCV 2026, University of Bologna).
Copyright (c) 2026 Fabio Tosi. Licensed under the MIT License.

## Modifications

LibreYOLO checkpoint-schema metadata wrap only. Learned parameters are
byte-identical to upstream. See `weights/convert_zipdepth_weights.py` in the
[LibreYOLO source repository](https://github.com/LibreYOLO/libreyolo).

## Training-data provenance

Upstream trained these weights by distilling pseudo-labels produced by Depth
Anything V2 Large (a CC-BY-NC-4.0 checkpoint) over ~14M images from 17 public
datasets. The MIT grant on the resulting student weights is the upstream
authors' published licensing position; the lineage is documented here for
transparency.

## License

MIT License. See the [`LICENSE`](./LICENSE) and [`NOTICE`](./NOTICE) files in
this repository.