Explore speech recognition model performance
Launch Argilla for data labeling and annotation
Select and analyze data subsets
Explore tradeoffs between privacy and fairness in machine learning models
Need to analyze data? Let a Llama-3.1 agent do it for you!
statistics analysis for linear regression
Generate detailed data reports
Generate a detailed dataset report
Detect bank fraud without revealing personal data
Display a Bokeh plot
Build, preprocess, and train machine learning models
Submit evaluations for speaker tagging and view leaderboard
View monthly arXiv download trends since 1994
WhisperKit Android Benchmarks is a data visualization tool designed to explore and analyze the performance of speech recognition models on Android devices. It provides detailed insights into how different models handle various audio inputs and conditions. Built for developers and researchers, this tool helps in optimizing speech-to-text systems and improving user experiences.
What devices are supported by WhisperKit Android Benchmarks?
WhisperKit Android Benchmarks is compatible with most Android devices running Android 8.0 (API level 26) or later.
Can I use custom audio files for benchmarking?
Yes, WhisperKit allows you to upload custom audio files for testing speech recognition models.
How do I interpret the benchmark results?
The app provides interactive visualizations and detailed metrics. Look for accuracy percentages, response times, and error rates to evaluate model performance.