Entity · paper

Multi-Expert Routing for Multi-Domain Low-Resource OCR: A Manchu Case Study

paperactiveprovisionalmulti-expert-routing-for-multi-domain-low-resource-ocr-a-manchu-case-study-09ad3378·1 events·first seen 5d ago

Aliases: Multi-Expert Routing for Multi-Domain Low-Resource OCR: A Manchu Case Study

More like this (12)

Redesign Mixture-of-Experts Routers with Manifold Power Iteration Azure OCR OCR-Robust DeepSeek-OCR-2 PP-OCRv6 A Multi-Domain Benchmark for Detecting AI-Generated Text-Rich Images from GPT-Image-2 When Does Combining Language Models Help? A Co-Failure Ceiling on Routing, Voting, and Mixture-of-Agents Across 67 Frontier Models How Robust is OCR-Reasoning? Evaluating OCR-Reasoning Robustness of Vision-Language Models under Visual Perturbations Reroute, Don't Remove: Recoverable Visual Token Routing for Vision-Language Models Multi-Task Learning Multilingual Coreference Resolution Shared Task Layer-Adaptive Expert Pruning

Recent events (1)

4arXiv · cs.AI·5d ago·source ↗

Multi-expert routing system achieves near-oracle OCR accuracy on low-resource Manchu historical scripts

A new arXiv preprint presents a multi-expert OCR system for historical Manchu manuscripts, routing pages by visual style (regular script, running script, chancery hand) using a lightweight image classifier. The router achieves 99.3% page-level domain accuracy and matches a domain-label oracle, with character error rates of 0.30%, 1.57%, and 4.83% across three script styles. The approach reuses fine-tuning checkpoints as domain specialists, avoiding the need to train dedicated experts for each domain. The work addresses low-resource OCR for endangered historical scripts with a reproducible evaluation protocol.

Multi-Expert Routing for Multi-Domain Low-Resource OCR: A Manchu Case Study