NJU-3DV/SpatialVID

• highreadme#1

  Add a clear, differentiating value proposition to the README's introduction

  Why:

  CURRENT

  The README's initial content focuses on the title and author list.

  COPY-PASTE FIX

  Copy

  SpatialVID is a pioneering large-scale video dataset specifically curated with rich spatial annotations to accelerate research in 3D generation, 4D reconstruction, and the development of advanced world models. It addresses the critical need for data that enables AI to understand and synthesize dynamic 3D scenes from video, going beyond typical autonomous driving or static scene datasets.

  OpenDoneDismiss
• mediumabout#2

  Expand the repository description to highlight key applications

  Why:

  CURRENT

  [CVPR 2026] SpatialVID: A Large-Scale Video Dataset with Spatial Annotations

  COPY-PASTE FIX

  Copy

  [CVPR 2026] SpatialVID: A large-scale video dataset with rich spatial annotations, specifically designed to advance research in 3D generation, 4D reconstruction, and the development of advanced world models.

  OpenDoneDismiss
• lowreadme#3

  Add a 'Key Use Cases' section to the README

  Why:

  COPY-PASTE FIX

  Copy

  ## Key Use Cases
  
  SpatialVID is uniquely designed to accelerate research and development in several cutting-edge areas:
  
  *   **3D Generation and 4D Reconstruction:** Leverage our rich spatial and temporal annotations to train models capable of generating realistic 3D scenes and reconstructing dynamic 4D environments from video input.
  *   **World Model Training:** Develop advanced AI world models that learn to predict future states and understand complex spatio-temporal dynamics by utilizing SpatialVID's comprehensive video data and spatial ground truth.
  *   **Vision-Language Model Development:** Enhance vision-language models with a deeper understanding of spatial relationships and object interactions within video contexts, enabling more sophisticated reasoning capabilities.

  OpenDoneDismiss

Category GEO backends resolved for this scan: google/gemini-2.5-flash, deepseek/deepseek-v4-flash

• CATEGORY QUERY

  Where can I find a large video dataset with spatial information for 3D generation?

  you: not recommended
  AI recommended (in order):

  1. Waymo Open Dataset
  2. nuScenes Dataset
  3. KITTI Vision Benchmark Suite
  4. Replica Dataset
  5. ScanNet
  6. Matterport3D
  7. Google Earth Studio

  AI recommended 7 alternatives but never named NJU-3DV/SpatialVID. This is the gap to close.

  Show full AI answer
• CATEGORY QUERY

  What are the best video datasets providing spatial annotations for training advanced world models?

  you: not recommended
  AI recommended (in order):

  1. Waymo Open Dataset
  2. nuScenes
  3. Argoverse 2 (Motion Forecasting Dataset)
  4. KITTI (Object Tracking and Odometry Benchmarks)
  5. BDD100K (Driving Scenes with Segmentation and Object Detection)
  6. Cityscapes (Video Sequences)
  7. MOT17/MOT20 (Multiple Object Tracking)

  AI recommended 7 alternatives but never named NJU-3DV/SpatialVID. This is the gap to close.

  Show full AI answer

• Metadata completeness

  pass

• README presence

  pass

• Compared to common alternatives in this category, what is the core differentiator of NJU-3DV/SpatialVID?

  pass
  ✓AI named NJU-3DV/SpatialVID explicitly

  AI answers can be confidently wrong. Read for accuracy: does it match your actual tech stack, audience, and differentiator?

• If a team adopts NJU-3DV/SpatialVID in production, what risks or prerequisites should they evaluate first?

  pass
  ✓AI named NJU-3DV/SpatialVID explicitly

  AI answers can be confidently wrong. Read for accuracy: does it match your actual tech stack, audience, and differentiator?

• In one sentence, what problem does the repo NJU-3DV/SpatialVID solve, and who is the primary audience?

  pass
  ✓AI named NJU-3DV/SpatialVID explicitly

  AI answers can be confidently wrong. Read for accuracy: does it match your actual tech stack, audience, and differentiator?