Speech Playground: An Interactive Tool for Speech Analysis and Comparison
2026-07-01 • Computation and Language
Computation and LanguageSound
AI summaryⓘ
The authors created Speech Playground, a tool that helps people see and compare different kinds of speech features easily. Unlike older software like Praat, their tool works well with new deep learning methods and lets users explore various speech data types interactively. It also supports important speech annotations and alignment options to help users listen and visualize speech comparisons. The tool is meant to help researchers and those working on pronunciation training experiments.
speech visualizationdeep learning representationsPraatTextGridforced alignmentspeech featurescomputer-aided pronunciation trainingspeech researchPython backendweb frontend
Authors
Stephen McIntosh, Daisuke Saito, Nobuaki Minematsu
Abstract
This paper presents Speech Playground, an interactive speech visualization and comparison tool. While existing tools such as Praat are excellent, it can be cumbersome to integrate them with modern deep learning representations and use them for comparison. Speech Playground addresses this by combining a Python backend with a web-based frontend for interactive exploration of multiple feature types, including continuous, discrete, and variable-length representations. It includes TextGrid and forced alignment support together with configurable distance and alignment settings for visual and auditory comparison. Speech Playground is intended for use in speech research, representation validation, and computer-aided pronunciation training (CAPT)-oriented experimentation.