In terms of computer vision and multimedia technologies, GTCOM has established a computer vision technology that can efficiently and automatically analyze the contents of massive image and video data, thus realizing the content-based intelligent identification, retrieval and analysis of multimedia information by deeply integrating resources with technologies and applying advanced technologies such as digital image processing, pattern recognition, deep learning and natural-language processing. GTCOM, working on the basis of high-quality video big data, has developed an array of video content analysis products, including video person identification, video classification, video semantic search and video content understanding, thus achieving in-depth excavation of the intrinsic value of video big data.
GTCOM's video-content analysis platform integrates core technologies such as timeline generation, intelligent speech analysis, image semantic analysis and video structured description. The platform uses video decoding and analysis to extract speech and images from videos, after which it converts video into text through the combination of technologies such as speech and image recognition. The platform can also find and locate specific texts in an original video. It analyzes speech spectra to recognize the endpoints of sentences and automatically generate the video timeline, thereby saving human and material resources. The platform can be widely used in the industry sectors of film, television and Internet video self-media, providing significant convenience while ensuring great commercial value for video producers.