Convert TSV to STM

Understanding tsv and stm file formats

TSV (Tab-Separated Values) is a simple text-based file format used for storing data in a tabular structure, where each field is separated by a tab character. It is commonly used for data exchange between spreadsheet applications and databases.

STM (Segment Time Marked) files are used primarily in speech and language processing. They contain time-aligned segment information for audio files, marking the start and end times of spoken segments along with speaker and transcription data. STM files are widely used in projects like the NIST Speech Recognition Evaluation.

Why convert tsv to stm?

Converting a TSV file to an STM file is often necessary when you have segment information in a tabular format (such as start time, end time, speaker, and transcription) and need to use it in speech processing tools that require the STM format.

How to convert tsv to stm

There is no direct, universal converter for TSV to STM since the conversion depends on the structure of your TSV file. However, you can use scripting languages like Python to automate the process. Here’s a general approach:

1. Open your TSV file in a spreadsheet editor or text editor to verify the column order (e.g., filename, channel, speaker, start time, end time, transcription).
2. Write a Python script to read the TSV file, parse each row, and format it according to the STM specification:
   filename channel speaker start_time end_time <NA> transcription
3. Save the output as a .stm file.

Recommended software and tools

• Python with pandas: Easily read TSV files and write custom scripts for STM formatting.
• LibreOffice Calc: For manual editing and exporting TSV data, though scripting is still needed for STM formatting.
• Custom scripts: For batch processing, Python or Perl scripts are the most flexible and reliable.

Example Python script for conversion

import pandas as pd

df = pd.read_csv('input.tsv', sep='\t')
with open('output.stm', 'w') as f:
    for _, row in df.iterrows():
        line = f"{row['filename']} 1 {row['speaker']} {row['start_time']} {row['end_time']} <NA> {row['transcription']}\n"
        f.write(line)

Adjust the column names as needed to match your TSV file.

Summary

Converting TSV to STM is best accomplished with a custom script, as the process depends on your data structure. Python is the recommended tool for this task due to its flexibility and ease of handling tabular data.