InterBEE REVIEW2016

54/71

Broadcasting equipment suppliers still rely on hardware for their income (54%), but it looks like software will soon exceed 50% of their income. Because broadcasting stations and the media industry are shifting to a software-lead busi-ness infrastructure using existing equipment, custom-made equipment manufacturers have been struggling. Existing sup-pliers, therefore, are shifting from dedicated hardware to soft-ware that runs on multi-purpose equipment.In the broadcasting markets of advanced nations, there is increasing investment in transmission-related products and services due to a shift to next-generation terrestrial standards (ATSC3.0 and DVB-T2).In emerging nations, however, there is still expected to be large industrial demand due to long-term shifts from SD to HD and analog to digital. For example, switchover to terrestrial digital broadcasting in the Asia-Pacific region is only com-plete in Japan and South Korea. As to the shift to UHD, it is not clear whether consumers want this. The development of devices, particularly mobile ones, is changing ideas about conventional broadcasting, so the increased efficiency of the industry is being pursued due to excess costs as well as to raise technological funds.Microsoft’s Azure Media Analytics is an AI/machine learn-ing cloud service developed to be easily used by people work-ing at broadcasting stations. It has several functions that gen-erate metadata for the user. One such function is to “convert spoken words to text”. The second is to identify certain people’s faces from images and assess emotions from eight types of feelings. The third, which is used for monitoring cam-eras and the like, isolates moving objects. The fourth extracts textual data that appears in moving images.Azure Media Analytics has another “clipping” function. That function automatically creates a clip of highlights, enabling about two hours of video to be smoothly condensed into about two minutes of highlights. Its last function is to blur certain people in an image.To give an example, when the image of a person swimming is imported into this service, not only can it identify that person’s sex, age and face, it will also automatically generate a sentence such as “this person appears to be swimming in a swimming pool”.84The IABM tracking the Broadcast and Media industry trends10:00 AM-10:40 AM, Thursday, November 17th Invited Session 1Mr. Peter BruceDirector, IABM APACDigital Transformation of the Media Business to the Cloud11:00 AM-11:40 AM, Thursday, November 17thInvited Session 2Mr. Tony EmersonManaging Director, Worldwide Media & Cable, Microsoft CorporationMr. Daiyu HatakeyamaTechnical Evangelist, Microsoft Japan Co., Ltd.