Performance Comparison on Common Scene Text Recognition Datasets (%)

A 9.5 KB Excel file containing performance metrics for the SVTRv2X model on multiple Scene Text Recognition (STR) benchmarks. The dataset was authored by Jian Guo and last updated on June 1, 2026. It likely contains tabular results comparing the proposed framework's recognition ability against other models.

Use Cases

• Benchmarking new STR models based on reported performance metrics.
• Analyzing model robustness to geometric distortions and complex backgrounds based on described evaluation scenarios.
• Comparing the efficiency of lightweight CTC-based models against enhanced architectures like SVTRv2X.
• Investigating model performance on diverse text styles as suggested by the integration of a Mixture-of-Experts module.

Strengths

• Dataset is small (9.5 KB), facilitating quick download and inspection.
• Released under a permissive CC-BY-4.0 license, allowing for broad reuse.
• Focuses on a specific, well-defined task of Scene Text Recognition benchmarking.

Limitations

• Row count and specific column definitions are unknown, limiting suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare

Collection Method

Likely contains experimental results from the author's research, as described.

Freshness

Last updated 2026-06-01 17:45:59; freshness should be verified.