Apple releases Depth Pro 2 monocular depth estimation method
Understand this story
Apple research team proposes Depth Pro 2, using a single scalable network to process both images and video and predict dense depth maps in meters at native resolution. Apple says the method simultaneously outperforms prior work in metric accuracy, sharpness, and temporal consistency, and has open-sourced the code and weights. Previous multi-frame methods emphasized temporal consistency while single-frame methods emphasized edge sharpness; Depth Pro 2 attempts to balance both. The above performance conclusions come from Apple's own statements.
AI overview based on source reports · Updated 54 min ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Apple Machine Learning ResearchDepth Pro 2: Sharp and Temporally Consistent Metric Monocular Depth
Depth Pro 2 is a single scalable network that estimates dense metric depth for images and video at native resolution, producing depth maps that are metrically accurate, sharp, and temporally consistent simultaneously. The authors say prior methods trade off temporal consistency against boundary sharpness, with multi-frame methods favoring consistency and single-frame methods favoring sharpness. They release a family of models defining a new Pareto frontier in metric accuracy, sharpness, temporal consistency, and runtime, along with code and weights.
Activity for this story
Not enough continuous observations to show a trend yet.