Ablation Experiment Results for Urban Road Crack Detection Method

5.5 KB of experimental results from a study proposing a novel road crack detection method. The dataset, authored by Jun Qiao and last updated in April 2026, contains results from an ablation experiment fusing U-net and ResNeXt networks. It likely contains performance metrics comparing the proposed method against others.

Use Cases

• Benchmarking new crack detection algorithms based on the described performance metrics.
• Analyzing the efficiency of memory usage in vision models based on the reported 2.1GB video memory peak.
• Training or validating models for horizontal and vertical crack classification based on the described detection targets.
• Studying model performance on block cracks and other crack types where the description notes room for improvement.

Strengths

• Results include a specific performance metric: a video memory peak controlled within 2.1GB.
• The dataset is openly licensed under CC-BY-4.0, facilitating reuse and sharing.
• The description provides a clear research context and comparison to other detection methods.

Limitations

• Row count and column-level documentation are unknown, requiring manual inspection after download.
• The dataset is very small at 5.5 KB, indicating limited scope, likely containing only summary results.
• The temporal and geographic coverage of the underlying image data is not specified.

Provenance

Source

Jun Qiao via figshare.

Collection Method

Likely derived from an ablation experiment comparing a proposed U-net and ResNeXt fusion method against other detection techniques.

Time Range

null

Freshness

Last updated 2026-04-21 17:34:40; freshness should be verified.

Geography

null