Class-wise mIoU Comparison on SensatUrban for Semantic Segmentation

A tabular dataset comparing class-wise mean Intersection over Union (mIoU) scores for semantic segmentation models on the SensatUrban benchmark. The data, stored in an XLS file of 5.5 KB, was uploaded by Zhe Jing in May 2026 under a CC-BY-4.0 license. It likely contains performance metrics for models like RPINet, which integrates 2D remote sensing images with 3D point cloud data.

Use Cases

• Benchmarking semantic segmentation models based on class-wise performance metrics.
• Analyzing the effectiveness of 2D and 3D feature fusion methods described in the paper.
• Comparing model performance on specific urban object classes within the SensatUrban dataset.

Strengths

• Explicitly states a state-of-the-art mIoU score of 66.5% for the proposed RPINet model.
• Provides a direct comparison metric (mIoU) for a specific, challenging urban dataset (SensatUrban).
• Released under a permissive CC-BY-4.0 license for reuse.

Limitations

• Dataset is very small (5.5 KB), suggesting limited scope, likely containing only summary metrics.
• Row count and column-level documentation are unknown; field semantics must be inferred after download.
• The description focuses on a research paper's results, not the dataset's internal structure.

Provenance

Source

figshare, associated with research on RPINet.

Collection Method

Likely derived from experimental results on the SensatUrban dataset.

Time Range

null

Freshness

Last updated 2026-05-22 17:38:41; freshness should be verified.

Geography

Based on the SensatUrban dataset, which covers urban areas.