Canonical: https://instain.ru/en/ai/depth-anything/
Language: en

Depth Anything Project  model

# Depth Anything. 

A depth map from a single photo

4.33 / 5    [Directory score](https://instain.ru/en/ai/depth-anything/#editorial-rating) [Official website ](https://depth-anything-v2.github.io/)

Original monocular depth estimation models. The developers publish relative depth models and separate metric variants.

![Depth Anything. The authors page with examples of Depth Anything V2 depth assessment.](https://instain.ru/assets/images/depth-anything-screen-5f4ec7bed2cf-1200.webp)

The authors page with examples of Depth Anything V2 depth assessment.  [Open the source ](https://github.com/DepthAnything/Depth-Anything-V2#depth-anything-v2)

## Relative positions of objects

Depth Anything builds a map that shows the estimated depth at the points of the shot. In relative mode, it is convenient to compare what is closer or further away. For scale in meters, authors publish individual trained options. Before taking the result into the project, specify the type of card and how to save it. A color scale image for viewing may contain less useful data than a numerical result.

## A window and a table in one frame

For a study sample, rent a room with a chair in the foreground and a window behind. Get a map and check whether the legs of the chair, table and far wall are different. Then look at the reflection in the window glass. The algorithm can evaluate it as part of a separate space, although a layer of glass is needed for the composition. If the card is used for weak camera movement, check what will appear in places that the chair has fenced.

Leave the original frame nearby. When the edge of the map is noticeably shifted, it is easier to compare it with a photo and manually limit the processing of the problem area.

## Choose model weights and a license

The repository describes the different scale of the model and individual options for metric estimation. Check available memory, specific weights license and the purpose of the selected version. To measure the real distance, you need control dimensions or an independent method of verification. In a visual composition, it is more important to make sure that the boundaries and relative positions of objects are suitable for the planned movement.

## Suitable tasks

- Prepare a depth map for the composition

- Check the visibility of the boundaries between plans

Features 

## Strengths

- Depth is estimated by one frame

- There are relative and separate metric options

Things to consider 

## Limitations

- A map can be wrong on transparent surfaces

- Relative depth does not give distance in meters

## Selection and conditions

Free access

Yes

Trial period

Not confirmed

No sign-up required

Not confirmed

Russian interface

Not confirmed

Local installation

Yes

API

Not confirmed

“Not confirmed” means there is not enough information. The feature may still be available; check the developer’s terms.

Input formats

Image

Result

Depth map

### Terms

For local processing, set the environment and download the appropriate weights. Developers document several sizes of models.

From task to first result

## Practical example

Task

Prepare a depth map for the composition

What to prepare

A picture with a clear foreground and several objects at different distances.

### Task brief or prompt

Build a depth map and compare the relative position of objects with the original image. Leave the source unchanged.

### Expected result

Depth map for visual verification.

### How to check the result

- Check near and far objects.

- Separately, look at the transparent surfaces and thin contours.

An editorial teaching example of a task brief, not the result of hands-on testing of the service. Check that the required mode is available, that you have rights to the source material and that the intended use of the result is permitted.

Check details

## Official sources

-  [Depth Anything V2 Project Page ](https://depth-anything-v2.github.io/)

-  [Code, weights and license terms Depth Anything ](https://github.com/DepthAnything/Depth-Anything-V2)

This description is based on the developer’s material. Features, prices and availability may change; check the primary source before use.

About the tool

Developer

Depth Anything Project

Category

[Images](https://instain.ru/en/category/images/)

Access

For local processing, set the environment and download the appropriate weights. Developers document several sizes of models.

Payment

Available without buying weight. Different options have different licenses and calculation requirements.

API

No separate API specified

[Developer terms ](https://github.com/DepthAnything/Depth-Anything-V2)

Directory score

4.33 / 5 

Features

4.50 / 5 

Workflows

4.10 / 5 

Ecosystem

4.40 / 5 

Documentation

4.30 / 5 

Scores are an editorial guide to the directory. They are not user reviews or the results of personal testing.

[See the methodology ](https://instain.ru/en/methodology/)

Keep exploring

## For similar tasks

[Full category  ](https://instain.ru/en/category/images/)

Images 

###  [Marigold](https://instain.ru/en/ai/marigold/)

Depth, surface normals, and albedo from one image

4.40 / 5   model  

Other result format: Image