Long-Short-term memory networks are artificial neural networks capable of learning other dependencies in time series or sequence predictions.
What is LSTM? You might have heard this term in the last interview you gave for a Machine Learning Engineer position, or some of your friends might have mentioned using LSTM in their predictive modeling projects. So the big question that may arise here is what LSTM is, what sort of projects can be created using the LSTM algorithm, etc. Do not worry. This article will cover the top 5 open-source projects in LSTM Neural Networks that everyone should know about in 2023.
Quick Fact: Microsoft uses LSTM models to improve its speech recognition software, selfies, search engine, learn to code, and more.
In today's article, we will talk about five open-source LSTM Neural Network projects/ repositories on GitHub to help you enhance your skills to prepare for the Data Science job market.
Note: In this article, we will talk about some excellent open-source LSTM neural network projects/Repositories that you can use in your projects in 2023. To read more about each, I recommend following the link given along the project.
If you are reading this article and have similar interests to me, I would suggest a few of my personal favourite courses and insist on completing this fantastic course by DataCamp.
Official Document: https://www.jaided.ai/
Easy OCR is a ready-to-use optical character recognition tool that supports 80+ languages, including their popular writing scripts. The languages supported include Latin, Chinese, Arabic, Devanagari, Cyrillic, etc.
The library can be installed using,
pip install easyocr and you can perform two tasks for development work. For extending the recognition model, read here, and for the detection mode (CRAFT), read here.
You can take a demo on their official website or use the link here.
The two main components of this project are:
- It can be operated for handwritten support.
- The code can be restructured to support swappable detection and recognition algorithms.
Stock Prediction models is a project comparison of machine learning and deep learning models that are majorly used for applying stock price forecasting. The project is also used for training trading bots and real-time simulation platforms.
There is a diverse list of LSTM models used in this project, some of which:
- LSTM Bidirectional
- LSTM 2-Path
- LSTM Seq2seq
- LSTM Bidirectional Seq2seq
- LSTM Seq2seq VAE
A standard jupyter notebook file allows you to forecast using any of the models mentioned above or including other models in the list.
The file for how-to-forecast can be found on this link.
LSTM human activity recognition implements real-time human activity detection using TensorFlow and an LSTM RNN network on a smartphone sensor dataset.
The type of movement is classified into six diverse categories. They are:
- Walking Upstairs
- Walking downstairs
The proposed method in the project uses the state-of-the-art LSTM model with an RNN model to avoid the feature engineering process, which generally requires a significant amount of time.
Automatic speech recognition is an end-to-end project for automatic speech identification for the English & Mandarin (Chinese) language. Some examples of generated speech predictions are:
it was about noon when captain waverley entered the straggling village or rather hamlet of tully veolan close to which was situated the mansion of the proprietor
it was about noon when captain wavraly entered the stragling bilagor of rather hamlent of tulevallon close to which wi situated the mantion of the propriater
one who writes of such an era labours under a troublesome disadvantage
one how rights of such an er a labours onder a troubles hom disadvantage
This is one of the finest projects for automatic speech recognition, as it is implemented in Tensorflow and supports training with CPU/GPU. It also gives you a choice to pick your network to train your model:
Official Document: https://arxiv.org/abs/1506.00019
LSTM repository is a basic implementation of the LSTM network in the python programming language in a few hundred lines of code from scratch. This repository is intended to keep the lstm concept simple by providing clean examples of training neural networks in lstm in python.
You can use this valuable yet helpful resource to learn how to implement an lstm network with minimum effort. You can extend your knowledge by adding more functional abilities and features and trying your LSTM model on different datasets.
The full article explaining the concept can be found here.
If you enjoy reading this article, we intercommunicate similar interests and are/will be in similar industries. So let’s connect via LinkedIn and Github. Please do not hesitate to send a contact request and check my PORTFOLIO!