D

Depth Anything Small Hf

Developed by Xenova
Small ONNX-format depth estimation model adapted for Transformers.js framework, suitable for web-based depth map prediction
Downloads 4,829
Release Time : 1/24/2024

Model Overview

This model is a compact ONNX version of Depth Anything, specifically designed for monocular depth estimation tasks, capable of predicting scene depth information from a single RGB image.

Model Features

Web adaptation
Direct browser execution via ONNX format and Transformers.js, eliminating server requirements
Lightweight design
Compact version reduces computational demands while maintaining accuracy
Real-time prediction
Optimized model enables near real-time depth map generation

Model Capabilities

Monocular depth estimation
Scene geometry understanding
Image depth map generation

Use Cases

Augmented Reality
AR scene understanding
Real-time scene depth estimation in browsers to support AR application development
Obtain scene geometry information without dedicated depth sensors
3D Reconstruction
Simple 3D modeling
Generate depth information from single photos to assist 3D model creation
Quickly obtain rough 3D structures of scenes
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase