Chronicles-OCR: A Cross-Temporal Benchmark for Chinese Character Evolution

Chronicles-OCR is the first benchmark designed to evaluate the cross-temporal visual perception capabilities of Visual Language Models across the complete evolutionary trajectory of Chinese characters, known as the 'Seven Chinese Scripts'. The dataset was curated in collaboration with institutional domain experts and is hosted on HuggingFace by author VirtualLUO. It was last updated on May 18, 2026.

Use Cases

• Benchmarking cross-temporal visual perception of VLLMs based on the 'Seven Chinese Scripts' concept.
• Training OCR models for historical Chinese documents based on the evolutionary trajectory of characters.
• Researching the visual evolution of written language based on the curated character samples.

Strengths

• Designed as the first comprehensive benchmark for cross-temporal perception of Chinese characters.
• Curated in collaboration with top-tier institutional domain experts.
• Covers the complete evolutionary trajectory known as the 'Seven Chinese Scripts'.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

VirtualLUO on HuggingFace, curated with institutional domain experts.

Collection Method

Likely involves collecting and annotating images of Chinese characters across historical scripts.

Freshness

Last updated 2026-05-18 08:39:41; freshness should be verified.