APARATUS FOR GENERATING A HIGHLIGHT VIDEO USING CHAT DATA AND AUDIO DATA

摘要

In the highlight image generation apparatus using chat data and audio data generated during the provision of a video, the extractor extracts a feature vector by separating any one of the chat data and the audio data into a predetermined time interval, the feature vector and the previous time of the extractor Forward learning unit that calculates the median value based on the result of learning the information of the interval in the recurrent neural network (RNN), and generates the cell state vector and the result vector of the corresponding time period based on the intermediate value, extraction The median value is calculated based on the result of learning the negative feature vector and the information of the subsequent time period in the Recurrent Neural Network (RNN), and the cell status vector and the result vector of the time period are calculated based on the median value. Highlight image generation apparatus using chat data and audio data, including a prediction unit that generates a score vector for a probability that a predetermined time period can be predicted as a highlight from the result vector of the backward learning unit and the forward generator and the result vector of the backward generator. Is disclosed.