Apple's New AI Model Creates 3D Depth Maps From 2D Images in Less Than a Second

Impressive work by Apple AI researchers.

PetaPixel
»#Apple releases #DepthPro, an #AImodel that is able to generate detailed #3Ddepthmaps from single #2Dimages in a fraction of a second — without relying on the camera data.« https://venturebeat.com/ai/apple-releases-depth-pro-an-ai-model-that-rewrites-the-rules-of-3d-vision/?eicker.news #tech #media
Apple releases Depth Pro, an AI model that rewrites the rules of 3D vision

Apple's Depth Pro AI model sets a new standard in 3D depth estimation, offering high-resolution, real-time depth mapping from a single image without camera metadata—transforming industries like AR, autonomous vehicles, and more.

VentureBeat

GitHub - apple/ml-depth-pro: Depth Pro: Sharp Monocular Metric Depth in Less Than a Second.

Link📌 Summary: 本文介紹了Depth Pro,一種能在不到一秒內提供清晰的單目公制深度估計的模型。Depth Pro能合成高解析度的深度圖,且具有絕對尺度,無需相機內部參數的數據。該模型在標準GPU上以0.3秒生成2.25百萬像素的深度圖,並包含多項技術貢獻,如高效的多尺度視覺變換器和專門的邊界準確性評估指標。用戶可安裝並運行模型以進行推斷,並可引用相關研究論文。

🎯 Key Points:
- Depth Pro模型實現了高分辨率的深度圖合成。
- 具備絕對尺度,無需額外的相機內部參數。
- 在標準GPU上運行迅速,每張2.25百萬像素深度圖僅需0.3秒。
- 采用多尺度變換器與真實及合成數據集結合訓練。
- 提供邊界準確度的專門評估指標。
- 用戶可透過簡單命令行或Python程式運行該模型。

🔖 Keywords: #DepthPro #深度估計 #單目 #高解析度 #邊界準確性

GitHub - apple/ml-depth-pro: Depth Pro: Sharp Monocular Metric Depth in Less Than a Second.

Depth Pro: Sharp Monocular Metric Depth in Less Than a Second. - apple/ml-depth-pro

GitHub
Apple’s AI research team just unleashed Depth Pro! This groundbreaking model generates a stunning 2.25-megapixel depth map in just 0.3 seconds on a standard GPU, using only a single image. 📸 Are we witnessing the future of imaging technology? #AI #Apple #TechInnovation #DepthPro

#開源分享 蘋果今天開源了一個深度學習項目:Depth Pro,不到一秒即可生成非常清晰和詳細的深度圖

可用於各種需要理解圖像深度的應用,利好自動駕駛汽車、虛擬實境、3D建模等場景

特點:
1、快速生成:可在0.3秒內生成一張225萬像素的深度圖
2、高清晰度:高解析度、高銳度、高頻細節,可以捕捉到很多細節
3、無需額外資訊:無需相機的特定設置資訊,比如焦距設置,就能生成準確的深度圖

專案地址: github.com/apple/ml-depth-pro

#DepthPro