Research map / 3D vision / 3D generation
Aria Digital Twin: A New Benchmark Dataset for Egocentric 3D Machine Perception
Xiaqing Pan, Nicholas Charron, Yongqian Yang, Scott Peters, Thomas Whelan, Chen Kong, Omkar Parkhi, Richard Newcombe, Yuheng (Carl) Ren
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| evaluate in office scene instead of apartment (lower ceilings for mocap cameras) | ADT ground-truth pipeline · ADT system accuracy | object translation error (mm): 6.94 -> 4.48 | Improved |
| evaluate in office scene instead of apartment | ADT ground-truth pipeline · ADT system accuracy | measured reprojection error (pixel): 6.9 -> 4.18 | Improved |
| optimize object pose against hand-labelled markers | ADT measured poses · ADT system accuracy | reprojection error (pixel): 6.78 -> 0.56 | Improved |
| use FPN instead of ViT-Det backbone | COCO-pretrained detectors on ADT · ADT 2D detection | AP-box (%): 11.42 -> 21.36 | Improved |
| use FPN instead of ViT-Det backbone | COCO-pretrained detectors on ADT · ADT 2D segmentation | AP-mask (%): 11.64 -> 19.81 | Improved |
| use Cube R-CNN instead of Total3D | 3D detectors on ADT · ADT 3D detection, chair | AP (%): 0.847 -> 3.72 | Improved |
| use Cube R-CNN instead of Total3D | 3D detectors on ADT · ADT 3D detection, fridge | AP (%): 0.048 -> 2.601 | Improved |
| use Cube R-CNN instead of Total3D | 3D detectors on ADT · ADT 3D detection, sofa | AP (%): 1.298 -> 1.252 | Hurt |
| apply LDM image translation to synthetic ADT images | Synthetic ADT renderings · ADT synthetic-to-real translation | PSNR: 16.383 -> 24.218 | Improved |
| apply LDM image translation to synthetic ADT images | Synthetic ADT renderings · ADT synthetic-to-real translation | LPIPS: 0.270 -> 0.126 | Improved |
| use Pix2Pix instead of LDM | Image translation on ADT · ADT synthetic-to-real translation | SSIM: 0.660 -> 0.674 | Improved |
| use TSIT instead of LDM | Image translation on ADT · ADT synthetic-to-real translation | PSNR: 24.218 -> 21.885 | Hurt |
Closest papers by meaning
- EgoSim: An Egocentric Multi-view Simulator and Real Dataset for Body-worn Cameras during Motion and Activity (NeurIPS 2024 · similarity 0.88)
- Seeing in the Dark: Benchmarking Egocentric 3D Vision with the Oxford Day-and-Night Dataset (NeurIPS 2025 · similarity 0.87)
- 3D Human Pose Perception from Egocentric Stereo Videos (CVPR 2024 · similarity 0.86)
- EgoGen: An Egocentric Synthetic Data Generator (CVPR 2024 · similarity 0.86)
- FRAME: Floor-aligned Representation for Avatar Motion from Egocentric Video (CVPR 2025 · similarity 0.86)
- EMHI: A Multimodal Egocentric Human Motion Dataset with HMD and Body-Worn IMUs (AAAI 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
