mgp-str-base

MGP-STR base-sized model is trained on MJSynth and SynthText. It was introduced in the paper Multi-Granularity Prediction for Scene Text Recognition and first released in this repository. MGP-STR is pure vision STR model, consisting of ViT and specially designed A^3 modules. The ViT module was initialized from the weights of DeiT-base, except the patch embedding model, due to the inconsistent input size. Images (32x128) are presented to the model as a sequence of fixed-size patches (resolution 4x4), which are linearly embedded. One also adds absolute position embeddings before feeding the sequence to the layers of the ViT module. Next, A^3 module selects a meaningful combination from the tokens of ViT output and integrates them into one output token corresponding to a specific character. Moreover, subword classification heads based on BPE A^3 module and WordPiece A^3 module are devised for subword predictions, so that the language information can be implicitly modeled. Finally, these multi-granularity predictions (character, subword and even word) are merged via a simple and effective fusion strategy.

翻訳準備中のため原文を表示しています