Depth Estimation Jobs
Computer vision roles requiring Depth Estimation expertise, across all industries and experience levels.
Open Positions
What is Depth Estimation?
Depth estimation recovers distance for each pixel, from stereo pairs, monocular images, or sequences. Monocular depth is geometrically ambiguous — scale cannot be recovered from a single image without priors — which is what makes learned approaches both powerful and unreliable in unfamiliar conditions.
Where Depth Estimation is used
Obstacle avoidance, occlusion handling for AR content placement, portrait-mode background separation on phones, and cheap depth sensing where LiDAR is not economical.
Roles that ask for Depth Estimation
- 3D Vision Engineer
- Perception Engineer
- Computer Vision Engineer
- AR/VR Engineer
- Research Scientist, Vision
Related skills & tools
Depth Estimation jobs — common questions
Why is monocular depth described as scale-ambiguous?
A small nearby object and a large distant one can project identically, so absolute scale is unrecoverable from one image alone. Models predict relative depth, and systems recover metric scale from a second sensor, known camera height, or object priors.
Which models are current?
MiDaS established robust relative monocular depth across domains; Depth Anything and metric-depth variants have pushed both generality and absolute accuracy. For stereo, RAFT-Stereo and its successors are the reference points.
Where does self-supervised depth fit in?
Training on video using photometric consistency between frames avoids the need for LiDAR ground truth. Monodepth2 popularised the approach, and it remains attractive for teams with abundant footage but no depth labels.