Quanta AI Labs

University of Toronto & Vector Institute · Depth Estimation

ProxyPose

Method for recovering full 6-DoF pose trajectories from monocular video via video-to-video translation, rendering a colored proxy cube that mirrors a query point's rigid motion and then applying contour fitting and PnP optimization; uses a LoRA-finetuned video generative model for temporal consistency.

Open the source ↗

Similar 3D & Spatial releases

  • PURE

    Peking University, HKUST, Pengcheng Laboratory · 3D Reconstruction · Week 44, 2026

  • GAE

    HKUST, ARC Lab (Tencent IEG), University of Hong Kong & UT Austin · Geometry-Native World Model · Week 40, 2026

  • Marigold V2

    HUAWEI Bayer Lab, EPFL & Bologna · Marigold V2 Depth · Week 37, 2026

  • Depth Map Control

    PixVerse · Depth Estimation · Week 31, 2026

  • ViDiHand

    ViDiHand · 3D Reconstruction · Week 27, 2026

  • Aleph Ultrasound Brain

    Aleph Neuro · 3D Reconstruction · Week 26, 2026

Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.