"The near future is online direct seeding is hot, because this has some of company,begin to cut direct seeding ancient bronze mirror to fizzle out this fractionize domain. For this, the following problem consulted net of thunder sharp edge to come from the graph general science and technology that specializes in pornographic identifying and the technologist in Yun Congke ability of computer vision company to undertake unscrambling.

At present company of many artificial intelligence already began Wu of Huang Ye of layout ancient bronze mirror, alibaba and Tecent are had respectively actor of green net, every phenomenon pursues in A yellow system of two big warning. And in poineering company side, relatively celebrated general having a plan is science and technology, flying search science and technology, deep graph wisdom is taken wait for a company.

The near future is online direct seeding is hot, because this has some of company,begin to cut direct seeding ancient bronze mirror to fizzle out this fractionize domain. For this, the following problem consulted net of thunder sharp edge to come from the graph general science and technology that specializes in pornographic identifying and the technologist in Yun Congke ability of computer vision company to undertake unscrambling.

The network matchs a plan

Direct seeding (like Chinese prickly ash, reflect a customer) , online video (if actor or actress cruel, love surprises art) , where is the distinction with picture yellow ancient bronze mirror? The yellow difficulty of ancient bronze mirror of direct seeding whether the biggest? What technology problem should solve?

Yun Congke ability is advanced and algorithmic engineer Zhou Xiang:

Actually these 3 (direct seeding, online video, picture ancient bronze mirror is yellow) difference is not big, 3 person but classify is handled for real time video and huge picture is handled.

Video / direct seeding is trends, the picture is a static state. When differentiating video and direct seeding, can turn dynamic content decipher into picture frame to judge, differentiate with static picture so do not have difference.

As to technical issue, ancient bronze mirror is actually yellow in algorithmic level difficulty not tall, after using deepness study algorithm to train, can achieve yellow result of right ancient bronze mirror.

Engineer of graph general science and technology is calculating a few small complement did on ability and algorithmic requirement:

Basically be to adopt removed cut at present pair of direct seeding of and so on of scheme of graph, crucial frame check and online video undertake discriminating handling, so final also be pair of sheet piece the processing of the picture.

But, as a result of the real time sex of direct seeding, so direct seeding photograph is mixed to video of other explicit line picture, the processing speed demand of the picture identifying to the machine is higher, this basically is to computation ability and algorithmic requirement are increased.