It’s an exciting time to be a developer in the voice computing space: 1 in 4 searches on Google are now voice-enabled, Amazon Alexa just passed 10,000 skills, and 100 million calls are completed on WhatsApp daily. But where do you go to start learning how to code in this field?
Whether you are a veteran developer or just starting out, this book guides you through the process of building voice-based applications in Python.
Applies >20 Python libraries to help you solve voice-related problems faster.
Get access to training voice datasets like the Common Voice Project or AudioSet.
Over 200 scripts are provided on GitHub to get you up-and-running quickly.
Become involved in the larger open source voice community, 45,000+ people and counting.
"I like the fact that there's a lot of code samples right after new concepts in the book are introduced. With this structure, I was able to quickly featurize and build machine learning models for voice datasets. The book is also filled with rich tables and visualizations; it's hard to find information like this in one place."
"I've been looking for a textbook for voice computing for several years now. In academic research I have struggled at times putting together material to efficiently orient new undergraduates, graduates, post-docs, and even other faculty to the field. A book like this is valuable to anyone doing work related to voice computing given the rapid growth in a brief period has outpaced the conventional channels to share this new knowledge. I will recommend this book to everyone I encounter who is aiming to do any work in voice computing."
"Collecting voice has many layers and challenges, Jim has successfully navigated the “do’s and dont’s” into an easy guide for starters and seasoned voice collectors. I highly recommend this book if you need to collect great quality of voice, but need those voice files to produce great data at all levels. This book accomplishes all of that and then some."
"This book provides an extraordinarily thorough overview of the entire field of speech technologies. Jim's practical ingenuity shines as he guides us through all the technical components and design choices involved in building a voice assistant. Microphone types, audio features for machine learning, a scalable backend via containerized microservices, security and ethical considerations, and much more, is all covered in this hands-on guide replete with working code. Use it to get your hands dirty in this exciting field — and even as a blueprint to create your own speech tech startup!"
Voice computing is the discipline that aims to develop hardware or software to process voice inputs. You can learn more about this field in the introductory video here.
The #Voicebook focuses on building voice computing software applications in Python.
This book is geared for beginners or veteran programmers alike. Many of the chapters start off simple then go into more advanced topics, so you're likely able to get something out of it regardless of your background.
Specifically, by the end of the book you'll be able to:
• Understand how to read/write, record, clean, encrypt, playback, transcode, transcribe, compress, publish, featurize, model, and visualize voice files
• Build your own voice computer and voice assistant from scratch
• Design cutting-edge microservice server architectures on top of Docker and Kubernetes
• Get access to 200+ starter scripts in a GitHub repository
• Become involved in the larger open source voice community
Simply, there are relatively few resources to learn how to write voice-enabled software that are straightforward and easy-to-understand.
Throughout the past 6 months, I have had repeated requests for resources to learn voice computing. There are a few great places to start - like sharing the documentation of a few modules (e.g. LibROSA); however, this is often not enough to get through the activation necessary to build good software. I would find myself taking some breakout sessions with Fellows to hack code together to model voice files. Soon, my time became limited as a CEO and I could not help many eager developers wanting to enter into this field.
This book is therefore a first attempt to scale this knowledge in a more repeatable way.
NeuroLex Laboratories is a company with the goal to make voice computing accessible to everyone. Over the past year, we have expanded our team to over 15 people, launched 7 research pilots, built a pipeline of over 80 startup and research collaborators, and have been featured in various press outlets like the Atlantic and PsychNews.
The Innovation Fellows Program is a competition to engage outstanding individuals with NeuroLex. Specifically, you propose a demo project alongside a mentor in one of three categories: research, data science, or software. You then execute this demo project over a 3 month period either individually or as part of a group. The program culminates in a Demo Day where you present your project to a panel of judges. Through this structure, you can gain hands-on expeirence and be part of a sustainable, lifelong community.
If you're interested, you can apply for the program @ innovate.neurolex.co
Please use this information:
Schwoebel, J. (2018). An Introduction to Voice Computing in Python. Boston; Seattle, Atlanta: NeuroLex Laboratories. https://github.com/jim-schwoebel/voicebook
Share this page!