EmbodiedOcc-ScanNet: 3D Occupancy Prediction Benchmark for Embodied Scene Understanding

EmbodiedOcc-ScanNet is a reorganized benchmark dataset for evaluating embodied 3D occupancy prediction tasks. It is based on local annotations from the ScanNet dataset and accompanies the paper 'EmbodiedOcc: Embodied 3D Occupancy Prediction for Vision-based Online Scene Understanding'. The dataset was created by YkiWu and was last updated on Hugging Face on 2025-08-27.

Use Cases

• Benchmarking embodied 3D occupancy prediction models based on local annotations.
• Training models for vision-based online scene understanding tasks.
• Evaluating the performance of algorithms for 3D semantic scene completion.
• Researching methods for real-time 3D environment perception for robotics or autonomous agents.

Strengths

• Dataset is explicitly designed as a benchmark for a specific research task (embodied 3D occupancy prediction).
• Accompanies a published academic paper, providing a research context.
• Last updated on 2025-08-27, indicating recent maintenance.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and dataset size are unknown, which may limit suitability assessment.
• Data may reflect the biases inherent to the original ScanNet source dataset.

Provenance

Source

YkiWu (author), based on annotations from the ScanNet dataset.

Collection Method

Reorganized benchmark based on local annotations.

Freshness

Last updated 2025-08-27 08:14:33; freshness should be verified.