Apple releases Depth Pro 2 metric monocular depth model
11/22/2026 — 11/22, 13:18·1 sources (all)·1 reports (all)
Key takeaways
- Apple published a paper titled Depth Pro 2: Sharp and Temporally Consistent Metric Monocular Depth on November 19. The model estimates true…
Background & analysis
According to a DEV Community post dated November 22, 2026 at 12:46, Apple published the paper "Depth Pro 2: Sharp and Temporally Consistent Metric Monocular Depth" (arXiv 2611.26603) on November 19, 2026, officially introducing the Depth Pro 2 model. The release was relayed by the community site rather than announced through Apple's own channels, so the underlying details rest on that single secondary account.
Depth Pro 2 is designed to estimate metric depth for every pixel at the input's original resolution. It accepts three kinds of input: a single still photo, an entire video, or a live camera feed. In each case the output is expressed in real metres rather than relative or normalized depth values, which means the model must infer absolute distance from visual cues rather than recording it directly.
The paper states that images and videos share one common network built on a multi-view Transformer paired with a DPT decoder. According to the authors, using this shared architecture reduces the depth flickering that tends to appear when a model processes video frame by frame, keeping the estimated depth temporally consistent across consecutive frames.
As of the available reporting, the only verifiable facts are the release date, the paper title and its arXiv identifier, the supported input types, the full-resolution pixel-level output, and the network structure. No further developments, benchmark results, or availability details have been reported yet, so the story currently stops at the paper's publication.
AI-generated from 1 reports · updated 25 minutes ago
- Heat index
- 98
- All sources
- 1
- All reports
- 1
- First seen
- 1 hour ago
Reports on this story headlines open the original
Apple published the paper "Depth Pro 2: Sharp and Temporally Consistent Metric Monocular Depth" on November 19, estimating metric depth for every pixel from a single photo, a whole video or a live camera feed at the input's full resolution, according to a DEV Community write-up. One multi-view transformer with a DPT decoder handles both images and video, which the paper says reduces depth flicker across frames.
DEV Community · AIAI score 52
Other stories people are talking about
- Heat index 1205RisingLean Community verifies Anthropic and Claude Riemann hypothesis proof22 sources
- Heat index 1048Anthropic Reaches $1.6 Trillion Market Value in First Trading Week After IPO21 sources
- Heat index 546RisingThinking Machines Lab Open-Sources Inkling 2 Mini with 32GB Model Card Recommendation10 sources
- Heat index 523Anthropic IPO total raise reaches $93.5 billion after greenshoe exercise18 sources
- Heat index 449SEC probes Nvidia accounting practices as shares post seven-week slide9 sources
How is heat calculated?About the methodHide
Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.
This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion

