Fuser Apps are here 🚀
Fuser Apps are here. Free generations for the next month 💫
Let's GoZoeDepth
byIntel Labs
Transform single photographs into spatially accurate metric depth maps with smooth physical volume and real-world distance gradients
How Zoe Depth Map works
Convert 2D source photography into calibrated depth maps through an automated metric workflow.
Upload your photograph
Provide a clear image with defined spatial layers and balanced lighting.
Predict physical metric depth
The metric bins architecture calculates true real-world spatial distances across every pixel.
Export smooth displacement maps
Download granular grayscale depth data ready for 3D environments, point clouds, or synthetic lens blur.
What Zoe Depth Map is good at
From dynamic metric binning to granular distance gradients, explore how ZoeDepth resolves true spatial volume.
Physical Metric Precision
Calculates real-world distances in meters using dynamic metric bins, preserving smooth spatial slopes and continuous gradients across indoor and outdoor environments.
Smooth Volumetric Transitions
Eliminates stepped cutoffs with continuous grayscale gradients from near foregrounds to distant horizons, ensuring artifact-free 3D displacement maps.
Zero-Shot Scene Generalization
Processes complex architectural layouts, outdoor terrain, and detailed portraiture without scene-specific fine-tuning or recalibration.
Seamless 3D Engine Integration
Produces distance-accurate depth plates ready for point cloud generation, camera parallax flythroughs, and realistic synthetic bokeh in downstream creative tools.
Made with Zoe Depth Map
Browse high-fidelity visual scenes demonstrating smooth depth gradients across varied creative disciplines.
Documentary reportage plate on textured geological formations
Commercial product hero frame with transparent materials
Album art composition with high-contrast color collision
Sculptural furniture showcase with precise material contours
High-fashion editorial spread with dynamic physical pose
What people build with Zoe Depth Map
See how 3D artists, visual effects teams, and photographers use metric depth maps across production pipelines.
Frequently Asked Questions
ZoeDepth predicts actual physical distance in meters rather than just relative foreground-to-background order. By combining the MiDaS backbone with a dynamic metric bins module, it generates continuous, volumetrically accurate depth gradients while maintaining zero-shot generalization across diverse environments.
ZoeDepth is ideal for 3D scene reconstruction, point cloud generation, precise displacement mapping, parallax camera moves, and synthetic depth-of-field blur. Its calibrated metric gradients provide smooth spatial volume without stepped edges, making it well-suited for export into 3D software like Blender, Unity, or Unreal Engine.
Avoid ZoeDepth when conditioning diffusion models for stylized generation, as relative depth models provide sharper, high-contrast silhouettes that ControlNet latches onto more effectively. It is also unsuitable for real-time or latency-critical applications due to heavier computational requirements, and it can miscalculate depth on mirrors, water, or transparent glass.
ZoeDepth was developed by Intel Labs and published on February 23, 2023.
High-contrast, well-lit photographs with clear depth separation between foreground and background yield the highest metric accuracy. For indoor scenes, keeping subjects within a 0.5 to 10-meter range leverages the model's metric precision. Avoid heavily compressed images with blocky compression artifacts or scenes dominated by heavy specular reflections.
Try Zoe Depth Map on Fuser
Transform single photographs into spatially accurate metric depth maps with smooth physical volume and real-world distance gradients