With the launch of affordable data plans, mobile users have started using the internet more than before. If you just consider YouTube as an example, you will get to know the way in which internet usage increased on daily basis.
YouTube says that in the last one year, watching videos on mobile grew by up to 400 percent. That is, around 80% of YouTube watching time comes from smartphones. David Powell, Director of Online Partnerships and Development, YouTube, APAC says,"We're seeing more hunger for YouTube content across all genres. In 2016, we saw over 500 creators across India get over 100k subscribers."
He also said that due to the significant rise in the number of viewers in India, they are revamping their effort to make India as one of their most powerful and successful content creator in the Asian community.
Along with this success, the Company also announced that its automatic captioning system can now detect other ambient sound effects like music, applause, and laughter and then displays it on the bottom of the screen. To enable this feature, the model known as Deep Neural Network (DNN) is applied here.
The company is trying further to expand this capability, so that it can differentiate between sounds and display captions like [Mild Applause] and [Raucous Applause]. The company has actually restricted those three common sounds for now because most of the video producers manually caption these sound effects currently.
Also Read: Top 5 YouTube Channels for movie lovers
Regarding this problem, Google engineer Sourish Chaudhuri says,"While the sound space is obviously far richer and provides even more contextually relevant information than these three classes, the semantic information conveyed by these sound effects in the caption track is relatively unambiguous, as opposed to sounds like [RING] which raises the question of "what was it that rang - a bell, an alarm, a phone?."
After this launch, user's worldwide are expecting some more exciting features to pop up in their favorite YouTube app. To know more about Deep Neural Network model and the algorithm used in this feature, you can visit Google's blog.