README.md

---
pipeline_tag: depth-estimation
license: apache-2.0
base_model: depth-anything/Depth-Anything-V2-Small-hf
library_name: zeromodels
tags:
- keras
- zeromodels
- depth-anything-v2
- depth-estimation
- arxiv:2406.09414
- pytorch
- jax
- tf
---

## ***See [our collection](https://huggingface.co/collections/zeromodels/depth-anything-v1-and-v2-6a8eaf5352197613b1655ac5) for all versions of Depth Anything V2.***

# Run Depth Anything V2 with Keras 3: JAX, PyTorch, or TensorFlow

[![GitHub](https://img.shields.io/badge/GitHub-ZeroModels-black?logo=github)](https://github.com/IMvision12/ZeroModels) [![Docs](https://img.shields.io/badge/Docs-Depth--Anything--V2-blue)](https://imvision12.github.io/ZeroModels/depth_anything_v2/) [![Collection](https://img.shields.io/badge/HF-Depth--Anything--V2%20collection-yellow)](https://huggingface.co/collections/zeromodels/depth-anything-v1-and-v2-6a8eaf5352197613b1655ac5)

# zeromodels/depth_anything_v2_small

Paper: [Depth Anything V2 (arXiv:2406.09414)](https://arxiv.org/abs/2406.09414) · [HF Papers](https://huggingface.co/papers/2406.09414)

Depth Anything V2 keeps V1's architecture and improves data (synthetic labels plus large-scale pseudo-labeling). Relative heads return unitless inverse depth; metric indoor/outdoor heads return metres.

For more details on the model, please go to the upstream [model card](https://huggingface.co/depth-anything/Depth-Anything-V2-Small-hf).

Pure-**Keras 3** conversion of [`depth-anything/Depth-Anything-V2-Small-hf`](https://huggingface.co/depth-anything/Depth-Anything-V2-Small-hf) for [zeromodels](https://github.com/IMvision12/ZeroModels). One implementation runs unmodified on **TensorFlow / Torch / JAX**.

This is a **monocular depth** checkpoint (`DepthAnythingV2DepthEstimation`, relative).

## ✨ Quick start

```python
import os
os.environ["KERAS_BACKEND"] = "torch"  # or "jax" / "tensorflow"

from PIL import Image
from zeromodels.models.depth_anything_v2 import DepthAnythingV2DepthEstimation, DepthAnythingV2ImageProcessor

model = DepthAnythingV2DepthEstimation.from_weights("zeromodels/depth_anything_v2_small")
processor = DepthAnythingV2ImageProcessor.from_weights("zeromodels/depth_anything_v2_small")

image = Image.open("your_image.jpg").convert("RGB")
output = model(processor(image)["pixel_values"], training=False)
depth = processor.post_process_depth_estimation(
    output, original_size=(image.height, image.width)
)
print(depth.shape)
```

Load any Depth Anything V2 variant the same way with `from_weights("zeromodels/<variant>")`:

| Variant | Hub | Output |
|---|---|---|
| `depth_anything_v2_small` | [`zeromodels/depth_anything_v2_small`](https://huggingface.co/zeromodels/depth_anything_v2_small) | relative |
| `depth_anything_v2_base` | [`zeromodels/depth_anything_v2_base`](https://huggingface.co/zeromodels/depth_anything_v2_base) | relative |
| `depth_anything_v2_large` | [`zeromodels/depth_anything_v2_large`](https://huggingface.co/zeromodels/depth_anything_v2_large) | relative |
| `depth_anything_v2_metric_indoor_small` | [`zeromodels/depth_anything_v2_metric_indoor_small`](https://huggingface.co/zeromodels/depth_anything_v2_metric_indoor_small) | metric indoor |
| `depth_anything_v2_metric_indoor_base` | [`zeromodels/depth_anything_v2_metric_indoor_base`](https://huggingface.co/zeromodels/depth_anything_v2_metric_indoor_base) | metric indoor |
| `depth_anything_v2_metric_indoor_large` | [`zeromodels/depth_anything_v2_metric_indoor_large`](https://huggingface.co/zeromodels/depth_anything_v2_metric_indoor_large) | metric indoor |
| `depth_anything_v2_metric_outdoor_small` | [`zeromodels/depth_anything_v2_metric_outdoor_small`](https://huggingface.co/zeromodels/depth_anything_v2_metric_outdoor_small) | metric outdoor |
| `depth_anything_v2_metric_outdoor_base` | [`zeromodels/depth_anything_v2_metric_outdoor_base`](https://huggingface.co/zeromodels/depth_anything_v2_metric_outdoor_base) | metric outdoor |
| `depth_anything_v2_metric_outdoor_large` | [`zeromodels/depth_anything_v2_metric_outdoor_large`](https://huggingface.co/zeromodels/depth_anything_v2_metric_outdoor_large) | metric outdoor |

## Tips

- Set `KERAS_BACKEND` **before** importing Keras / zeromodels.
- Indoor and outdoor metric heads are not interchangeable.
- See [Depth Anything V2 docs]({DOCS_URL}) and [Loading Weights](https://imvision12.github.io/ZeroModels/loading_weights/).
- Community / upstream weights: `DepthAnythingV2DepthEstimation.from_weights("hf:depth-anything/Depth-Anything-V2-Small-hf")`.

## Special Thanks

A huge thank you to the Depth Anything authors for creating and releasing these models.

License: Apache 2.0.