Sequence-to-Label Script Identification for Multilingual OCR

We describe a novel line-level script identification method. Previous work repurposed an OCR model generating per-character script codes, counted to obtain line-level script identification. This has two shortcomings. First, as a sequence-to-sequence model it is more complex than necessary for the sequence-to-label problem of line script identification. This makes it harder to train and inefficient to run. Second, the counting heuristic may be suboptimal compared to a learned model. Therefore we reframe line script identification as a sequence-to-label problem and solve it using two components, trained end-toend: Encoder and Summarizer. The encoder converts a line image into a feature sequence. The summarizer aggregates the sequence to classify the line. We test various summarizers with identical inception-style convolutional networks as encoders. Experiments on scanned books and photos containing 232 languages in 30 scripts show 16% reduction of script identification error rate compared to the baseline. This improved script identification reduces the character error rate attributable to script misidentification by 33%.