Handwritten Text Recognition Lives in the High-Pixel Variance Subspace

The paper investigates self‑supervised pretraining for Handwritten Text Recognition (HTR) and finds that pixel‑reconstruction methods outperform contrastive methods, a contrast to natural‑image classification. The authors argue that discriminative information for HTR is concentrated in high‑variance pixel directions, and they test six SSL methods across six handwriting benchmarks in five languages, showing that pixel‑grounded approaches achieve the lowest character error rates and are more label‑efficient. When combined with a pretrained LLM decoder, a frozen pixel‑grounded encoder rivals fully fine‑tuned supervised baselines, underscoring the importance of aligning representations with high‑variance pixel subspaces.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

• llms
• computer-vision
• fine-tuning
• benchmarks
• safety