Day 10: Transcription with 3 lines of Python

Amazon Transcribe, Google Speech-to-Text, Azure Cognitive Services, IBM Watson, AssemblyAI, DeepGram, Speechmatics, and Rev, …all provide APIs to transcribe audio files. So why should you care about Picovoice Leopard? Just for the Free Tier? Nope! We have more reasons:

• Private: Voice data is processed on-device
• Accurate: Backed by an Open-Source Benchmark
• Compact and Computationally Efficient
• Cross-Platform: Runs on Linux, macOS, Windows, Android, iOS, Raspberry Pi, and NVIDIA Jetson
• Hyper-customizable: Self-service UI to customize models
• You can get started with 3 lines of code!

Let's get started!

1- Install
Install Leopard from a terminal:

pip3 install pvleoparddemo

2- Try it
Grab your free AccessKey from Picovoice Console and run the microphone demo from the terminal:

leopard_demo_mic --access_key ${YOUR_ACCESS_KEY}

3- Build
Create an instance of Leopard:

from pvleopard import *
o = create(access_key=${YOUR_ACCESS_KEY})

Transcribe an audio file:

transcript, words = o.process_file(${YOUR_AUDIO_FILE_PATH})
print(transcript)

Enjoy!