Developing and understanding Automatic Speech Recognition (ASR) systems is an inter-disciplinary activity, taking expertise in linguistics, computer science, mathematics, and electrical engineering.
When a human speaks a word,they causetheir voice to make a time-varying pattern of sounds. These sounds are waves of pressure that propagate through the air. The sounds are captured by a sensor, such as a microphone or microphone array, and turned into a sequence of numbers representing the pressure change over time. The automatic speech recognition system converts this time-pressure signal into a time-frequency-energy signal. It has been trained on a curated set of labeled speech sounds, and labels the sounds it is presented with. These acoustic labels are combined with a model of word pronunciation and a model of word sequences, to create a textual representation of what was said.
Instead of exploring one part of this process deeply, this course is designed to give an overview of the components of a modern ASR system. In each lecture, we describe a component's purpose and general structure. In each lab, the student creates a functioning block of the system. At the end of the course, we will have built a speech recognition system almost entirely out of Python code.
edX offers financial assistance for learners who want to earn Verified Certificates but who may not be able to pay the fee. To apply for financial assistance, enroll in the course, then follow this link to complete an application for assistance.
Note: These courses will retire in June. Please enroll only if you are able to finish your coursework in time.
Berbelek is taking this course right now, spending 1 hours a week on it and found the course difficulty to be hard.
My review is based on material of the first module only. The course seems to be strongly text-based, with no satisfying clarification of some terms and ideas. I believe video would be perfect to explain the subject - and the only thing we've got is a wall of text of mediocre clarity and quality. I expected nice introduction to the subject, but think nobody without prior knowledge would be able to finish the course.
Anonymous is taking this course right now.
The content of the course is very poor. Very complex concepts are captured in a form of brief articles which will not be helpful to anyone who does not already have a good grasp of them. The tutors only appear in the short introductory videos. I cannot...
The content of the course is very poor. Very complex concepts are captured in a form of brief articles which will not be helpful to anyone who does not already have a good grasp of them. The tutors only appear in the short introductory videos. I cannot comment on the labs and assignments because I decided not to pay for the course after realising how poor its content is. Perhaps there is some value in the fact that you can verify your implementations using a grader but I wouldn't count on it. Most importantly a lot of the course content is dedicated to the outdated solutions. Deep learning based solutions are also very briefly described and their implementations use CNTK framework rather than Tensorflow or some other popular machine learning framework. To anyone interested in learning about state-of-the-art speech recognition solutions I would rather recommend having a look at nVidia's OpenSeq2Seq toolkit.
Anonymous completed this course.
It is very text heavy course with few videos here and there explaining the module. Seems like the instructors are reading the script which is very obvious in the recordings as well as makes the learning experience very quirky.
The course is designed based on the assumption that you already have some understanding of the formulas and what they are talking.
In the past, I've read through materials and learned a lot. But this courses text material is too much of a read and doesn't do any justice explaining concepts in simple language.