Update README.md

README.md

@@ -1,2 +1,18 @@
-# Easter2
-Easter2.0: IMPROVING CONVOLUTIONAL MODELS FOR  HANDWRITTEN TEXT RECOGNITION
+# Easter2.0: IMPROVING CONVOLUTIONAL MODELS FOR  HANDWRITTEN TEXT RECOGNITION
+
+This repo provides the model and code for our paper: To Be Updated.
+
+[[PDF]](To Be Updated.)
+
+### Overview
+In this paper, we proposed a convolutional architecture for the task of handwritten text recognition that utilizes only 1D
+convolutions, dense residual connections and a SE module. We also proposed a simple and effective data augmentation
+technique-T ACo useful for OCR/HTR tasks. We have presented experimental study on components of Easter2.0
+architecture including dense residual connections, normalization choices, SE module, TACo variations and few-shot
+training. Our work achieves SOTA results on IAM-Test set when training data is limited, also Easter2.0 has very
+small number of trainable parameters compared to other solutions. The proposed architecture can be used in search of
+smaller, faster and efficient OCR/HTR solutions when available annotated data is limited.
+
+## Citation
+If you find our work helpful, please cite the following:
+To be updated.