Based on SmarCo’s audio and video one machine, the use of deep learning video content to understand technology, output video category, video content tags, and automatically select a representative video cover, to achieve video content comparison and video content review
1. Video classification based on the comprehensive analysis of the video voice and image, generating a category tag for video content.
2.The video fine-grained identification for the vertical field of individual modeling, fine identification and labeling of the video appeared in the object type, model and other details of the characteristics and output and the semantic understanding of the video.
Video classification application
1. Video content tag extraction
Through the video content analysis, quickly generated for the video commonly used tags, can be used to solve problem which the video can’t be retrieved because of lack of keywords.
2. Video recommended
Through the video content analysis technology, automatically extract the video content tags, to solve the video cold start recommended problem, increase the video ratings.
Based on the SmarCo’s audio and video one machine, the use of deep learning to identify pictures and videos in the Jurisprudence,、involving violence,、fear、political sensitivity, micro-advertising and other content, but also quickly and accurately from the beautiful and clear and other dimensions of the image video Screening, liberation of manpower
Using the depth of learning technology, the semantic analysis of the video, according to the user business were given sensitive, questioned, the normal three confidence scores, and then use the confidence score on the illegal video automatically filter, timely removal of illegal、fuzzy and other unqualified video
Combined with the SmarCo for the artificial intelligence, large data processing and other optimized design of the new high-throughput features of the server products, SmarCo large data one machine, through the combination of hardware and software to enhance the processing capacity, high throughput, high-strength real-time to ensure the integration of high-performance features, achieving mass video data training、analysis,、modeling, to provide intelligent video image analysis and content audit solution
l Technology leader. Provide industry-leading voice, image, natural language and other artificial intelligence technology, can be deep understanding of mining video content implied information for the application to provide data base.
l Application coverage. Full coverage from the video storage, codec, distribution and other basic capabilities to video content intelligent analysis capabilities and intelligent application capabilities.