Speech Recognition using KALDI

Speech Recognition using KALDI

Speech Recognition using KALDI   The people who are searching and new to the speech recognition models it is very great place to learn the open source tool KALDI. It is a open source tool kit and deals with the speech data. And the KALDI is mainly used for speech recognition, speaker diarisation and speaker recognition.   KALDI , it is mainly written in c/c++ and it is cover with the bash and python scripts. The wrapping spares are used to get into the deep source code. Before this, we have to know the available open source speech recognition tools with…

Deep Dive into Bidirectional LSTM

Deep Dive into Bidirectional LSTM

Deep Dive into Bidirectional LSTM Bidirectional LSTMs are an extension to typical LSTMs that can enhance performance of the model on sequence classification problems. Where all time steps of the input sequence are available, Bi-LSTMs train two LSTMs instead of one LSTMs on the input sequence. The first on the input sequence as-is and the other on a reversed copy of the input sequence. By this additional context is added to network and results are faster.   Bidirectional LSTMs The idea behind Bidirectional Recurrent Neural Networks (RNNs) is very straightforward. Which involves replicating the first recurrent layer in the network…

Maxima vs Minima and Global vs Local in Machine learning

Maxima vs Minima and Global vs Local in Machine learning

Maxima vs Minima and Global vs Local in Machine learning   You might have heard or read the statement that goes something like “The algorithm might get stuck at one of the local minima and not converge to the global minimum”. But what exactly does it mean? I’ll explain the concept of maxima, minima, local and global. Additionally I’ll be covering how calculus makes it possible to identify these points. I’ll explain the concepts for functions of single variable because they are easy to visualize. However, they extend to multivariate cases. Let us start with a few definitions.   Global Maximum:…