Thursday, June 13, 2013

Manual Transcription – A Thing of the Past?

Voice recognition software has become quite a hot topic in the recent past and has made quite a few developments and advancements, but the real question on everyone’s mind (Even ours!) is, “Can Voice recognition software really replace manual transcription?” Well, many people have many opinions about it, but we think it’s safe to assume that although voice recognition is being used for many purposes, it won’t be taking over the world as we know it any time soon.

What is Voice Recognition Software?              

We’ve spoken about this before, but for our new readers voice recognition software can simply be defined as an application that permits the speaker to speak into the computer and the computer takes it as a command (E.g.: Saving a file). During the last decade voice recognition (aka speech recognition) has been praised as the new and correct way of dictating information to the computer, but we all know that improvements need to be made. There are software’s out there that claim to have a 99% rate of accuracy when it comes dictation and also its three times faster than typing on a keyboard.

Transcription Only?

Now, what if you want to use this software for an interview? Can it be done? Well, in our opinion that is not the case! We’ve spoken before about voice recognition software being able to recognize voices, well same rule applies here. If, for example, there are three people at the interview and all three are speaking, the voice recognition software will only recognize the voice that it has been trained on and that can be a major setback for this software. If you want to record an interview the software will most likely only record your voice, however recording the interviewees voice will be much difficult and in most cases impossible. But there is another option if you’re willing to work a bit harder. You could record the interview on tape and repeat the words which the interviewee said at the interview into the voice recognition software. This way both responses will be saved but with the use of only one voice. But who has time for that?! Anybody would want to use this software at once and not repeat the whole conversation all over again just because your voice recognition software is trained to understand only one voice.

What’s Stopping this Software from Taking Over the World?

A seminar conducted by an expert on Digital Dictation argued that this software would only work for three in ten people. Why? Well the answer is simple; the software needs to familiarize itself with your voice, a process which would normally take under a few minutes, but here’s the catch! The computer will require much more time to be trained and be able to recognize words that it is not familiar with, or cannot understand because of your pattern of speech. This is a continuous process and therefore takes time.

Present software, when compared to the products available about five years ago, has come a long way! But can there be improvements made? Of course! As with any software out there in the market today developments can always be made. There’s no software or product that cannot be improved! Same goes for Voice Recognition, but what we must focus on is that this software is here! It’s already made quite a splash on the world and soon we will be able to test this software on almost every device that we own. The software might be requiring some adjustments, but look at the positive impact that it has already had on the world. We have mobile phones and even televisions that are operated with this software! Everyone (even without their knowledge) is using voice recognition software or at least possesses a device that is voice recognition enabled. Think of someone that is completely handicapped when it comes to operating an electronic device. If learning how to use something is difficult they can now simply use their voice to command the computer/device and in mere seconds it’s done!

There are people that use voice recognition software occasionally to give their hands and fingers a break from all the typing. If you’re used to manually typing, then afterwards you will need to go through the document and proofread it to make sure that there are no errors. When using the software you could maybe follow a process where you could read the words out and also check for any mistakes that need to be cleared. This can also save you quite a bit of time since you will be multi tasking.

Speech recognition software is not intelligent enough to know which homonyms to use for which has been dramatically improved in recent years. Some software helps with this problem to a certain extent. Words like “here” and “hear” or “two” or “too” still need to be defined as separate words, but in shorter sentences this problem can be solved. When considering a longer sentence the software will usually struggle to find the correct words for the correct meaning and this is where software developers need to pay more attention.

Another important fact is that when you’re typing on a keyboard you always have the option of giving commands to make the words bold, underline them, etc. However if you want to do this voice recognition software that also will need to be done in the form of a command, but question still remain. How could we command the computer to bold or underline specific words? Or even sentences? A small pause between the end of the dictation and the beginning of the command could be one way of doing things, but again this also slows down the whole process. Going back to re-speaking an interview could also be a tedious task. Each time a different participant starts to speak an indication needs to be made that there is another speaker about to speak.

Sufficient to say that we have come a long way since the first introduction of this software and that’s great! People are constantly looking for ways to get more things done in a shorter amount of time. Speech Recognition is one such way that people are getting more things done with less effort. Can there be improvements made? YES! But here’s the deal. Talking computers, phones and televisions were considered the devil’s work decades ago! But here we are utilizing them every single day. So it’s our opinion that given a few more years this software can really make an impression on the world and will be one of those things that people can’t live without!