Using sentence-transformer for symmetric semantic search — The topic for today is about calculating the similarity score between two sentences of the same or different languages. We will be utilizing the sentence-transformer framework which comes with its own pre-trained multilingual transformer models. We can make use of these models to compute the text embedding for more than…

Moving doesn’t always have to be hard — After reading this piece, you will learn to package your Docker images and deploy them on other machines locally. The process for copying a Docker image can be confusing for most developers as Docker comes with different commands for saving/loading and exporting/importing. …

Simple phonemization of words and texts in many languages — By reading this piece, you will learn how to convert an input text string to its corresponding phonemes in Python. A phoneme represents the smallest unit of sound in a language. For example, the word tab consists of three phonemes: /t/ /a/ /b/ The element b is distinguishable when compared…

A comprehensive open-source multi-lingual speech data — Speech synthesis, also known as text-to-speech (TTS) is one of the new key technologies in the artificial intelligence domain. It provides the capabilities to generate human-like voices from text input dynamically. TTS can be applied in a variety of purposes and tied closely with automation services. However, training a text-to-speech…

Tips and tricks to preprocess image datasets — This tutorial covers the following step-by-step guides: convert XML annotations to YOLO annotations visualize the bounding boxes in image using the newly created YOLO annotations split the datasets into train, validation and test sets Overview PASCAL VOC XML The PASCAL Visual Object Classes (VOC) project is one of the earliest computer vision project that…

All-in-one augmentation packages for machine learning — In the world of machine learning, data augmentation is one of the most useful techniques to enhance the performance of ML models. Data augmentation serves to create synthetic data via slight modification or transformation to the existing data. This helps to: increase the amount of training and test data. reduce…

Removing Personally Identifiable Information (PII) from transcription text In my previous article, I have covered How to Transcribe Audio Files to Text. In this tutorial, let’s explore a little further on how to remove Personally Identifiable Information (PII) from the transcription. …

Replace offensive words with asterisks Previously, I have covered a tutorial on Speech Content Safety Detection, which identifies sensitive content such as pornography, terrorism, and hate speech in speech. This feature is extremely useful for content moderation on platforms but it is over-excessive and over-egging the pudding if your use…

Identify relevant topics based on IAB Taxonomy Topic detection is a technique to discover the abstract topic behind a collection of documents. It is mostly part of the natural language processing technique to classify text into specific topics/domains. Although there are no rules and regulations on how topics should be…