Hi, I am a ML/DL expert, currently working as a deep learning specialist researcher at a mathematics research center.
I have rich experience in computer vision, using CNNs, for both practival and research purposes. I can accomplish the task of character classification/recognition using Python and TensorFlow/Keras, with a residual deep convolutional network. With a quality dataset 95%+ accuracy should be achievable.
I am using OpenCV and Pillow in Python for data preprocessing, dataset creation and augmentation.
If there aren't enough data, I can help with building a dataset based on some examples, finding data from the internet and/or using data augmentation.
I create TFLite-compatible models with TensorFlow, so inference can be ran on low-resource PCs, mobile and embedded devices. Deployment is possible with Python scripts / C++ or Java environments.