This repository lets you train neural networks models for performing end-to-end full-page handwriting recognition using the Apache MXNet deep learning frameworks on the IAM Dataset.
Hi.. i've executed "2_line_word_segmentation.ipynb" .. I am doing my undergrad thesis on "Bangla handwritten word detection" .. i've used this code as a baseline . now how can i improve its detection rate and predict proper bounding box? i've also added around 100 bangla images . If i am to make it SSD500 what are the changes i can make in this current code .. i'm using colab . and i've studied several blogs and paper too . but i'm not getting any proper solution . i've to submit my report within 20 jan. It will be a great help if someone could help me with these questions ..
Hi.. i've executed "2_line_word_segmentation.ipynb" .. I am doing my undergrad thesis on "Bangla handwritten word detection" .. i've used this code as a baseline . now how can i improve its detection rate and predict proper bounding box? i've also added around 100 bangla images . If i am to make it SSD500 what are the changes i can make in this current code .. i'm using colab . and i've studied several blogs and paper too . but i'm not getting any proper solution . i've to submit my report within 20 jan. It will be a great help if someone could help me with these questions ..