Instructions to use LibreYOLO/LibreZipDepthb-depth with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- LibreYOLO
How to use LibreYOLO/LibreZipDepthb-depth 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
Initial upload: ZipDepth b (MIT, metadata wrap of upstream checkpoint)
Browse files- LICENSE +21 -0
- LibreZipDepthb-depth.pt +3 -0
- NOTICE +12 -0
- README.md +57 -0
LICENSE
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
MIT License
|
| 2 |
+
|
| 3 |
+
Copyright (c) 2026 Fabio Tosi
|
| 4 |
+
|
| 5 |
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
| 6 |
+
of this software and associated documentation files (the "Software"), to deal
|
| 7 |
+
in the Software without restriction, including without limitation the rights
|
| 8 |
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
| 9 |
+
copies of the Software, and to permit persons to whom the Software is
|
| 10 |
+
furnished to do so, subject to the following conditions:
|
| 11 |
+
|
| 12 |
+
The above copyright notice and this permission notice shall be included in all
|
| 13 |
+
copies or substantial portions of the Software.
|
| 14 |
+
|
| 15 |
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
| 16 |
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
| 17 |
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
| 18 |
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
| 19 |
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
| 20 |
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
| 21 |
+
SOFTWARE.
|
LibreZipDepthb-depth.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:069ebb29c136b3e4b3613fd73dadf1f83d61fa4ee4312fc601b923f12f34fe67
|
| 3 |
+
size 27302611
|
NOTICE
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
LibreZipDepth weights
|
| 2 |
+
---------------------
|
| 3 |
+
|
| 4 |
+
This product contains weights derived from ZipDepth
|
| 5 |
+
(https://github.com/fabiotosi92/ZipDepth).
|
| 6 |
+
Copyright (c) 2026 Fabio Tosi.
|
| 7 |
+
Licensed under the MIT License.
|
| 8 |
+
|
| 9 |
+
The weights were trained by the upstream authors via knowledge distillation
|
| 10 |
+
from Depth Anything V2 Large pseudo-labels (~14M images across 17 public
|
| 11 |
+
datasets). The MIT license on these student weights is the upstream authors'
|
| 12 |
+
published position; this notice records the distillation lineage.
|
README.md
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
library_name: libreyolo
|
| 4 |
+
pipeline_tag: depth-estimation
|
| 5 |
+
tags:
|
| 6 |
+
- depth-estimation
|
| 7 |
+
- monocular-depth
|
| 8 |
+
- zipdepth
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# LibreZipDepthb-depth
|
| 12 |
+
|
| 13 |
+
ZipDepth base: a ~6M-parameter reparameterizable CNN (RepVGG encoder, SPPF +
|
| 14 |
+
cross-scale neck, FPN decoder with convex upsampling) for zero-shot relative
|
| 15 |
+
monocular depth estimation, repackaged for LibreYOLO. This is the standard
|
| 16 |
+
GPU/CPU checkpoint; the sibling repo
|
| 17 |
+
[LibreZipDepthbnpu-depth](https://huggingface.co/LibreYOLO/LibreZipDepthbnpu-depth)
|
| 18 |
+
carries the separately trained unfold-free decoder for NPU/edge compilers.
|
| 19 |
+
|
| 20 |
+
## Usage
|
| 21 |
+
|
| 22 |
+
```python
|
| 23 |
+
from libreyolo import LibreYOLO
|
| 24 |
+
|
| 25 |
+
model = LibreYOLO("LibreZipDepthb-depth.pt") # auto-downloads from this repo
|
| 26 |
+
results = model.predict("image.jpg", save=True)
|
| 27 |
+
depth = results[0].depth_map.data # (H, W) relative inverse depth
|
| 28 |
+
```
|
| 29 |
+
|
| 30 |
+
Outputs follow LibreYOLO's depth task contract: a dense float map on the
|
| 31 |
+
original image canvas, higher values mean closer to the camera, no metric unit.
|
| 32 |
+
|
| 33 |
+
## Source
|
| 34 |
+
|
| 35 |
+
Derived from [fabiotosi92/ZipDepth](https://github.com/fabiotosi92/ZipDepth)
|
| 36 |
+
(`checkpoints/zipdepth_base.pth`) at commit
|
| 37 |
+
`6b96f4d205f8a2e5377e81c1b74cc99a47f6693a` (ECCV 2026, University of Bologna).
|
| 38 |
+
Copyright (c) 2026 Fabio Tosi. Licensed under the MIT License.
|
| 39 |
+
|
| 40 |
+
## Modifications
|
| 41 |
+
|
| 42 |
+
LibreYOLO checkpoint-schema metadata wrap only. Learned parameters are
|
| 43 |
+
byte-identical to upstream. See `weights/convert_zipdepth_weights.py` in the
|
| 44 |
+
[LibreYOLO source repository](https://github.com/LibreYOLO/libreyolo).
|
| 45 |
+
|
| 46 |
+
## Training-data provenance
|
| 47 |
+
|
| 48 |
+
Upstream trained these weights by distilling pseudo-labels produced by Depth
|
| 49 |
+
Anything V2 Large (a CC-BY-NC-4.0 checkpoint) over ~14M images from 17 public
|
| 50 |
+
datasets. The MIT grant on the resulting student weights is the upstream
|
| 51 |
+
authors' published licensing position; the lineage is documented here for
|
| 52 |
+
transparency.
|
| 53 |
+
|
| 54 |
+
## License
|
| 55 |
+
|
| 56 |
+
MIT License. See the [`LICENSE`](./LICENSE) and [`NOTICE`](./NOTICE) files in
|
| 57 |
+
this repository.
|