A video processing apparatus includes a camera to continuously capture an image of an object to acquire video data, a memory, and circuitry to identify, from among a plurality of users appearing in the video data, a user who is speaking at a point in time when the video data is acquired as a currently-speaking user, store, in the memory, speech history information that associates, for each point in time when the video data is acquired during at least a predetermined time period, the currently-speaking user with time information indicating the point in time when the video data is acquired, and based on the speech history information, identify a first user currently speaking and a second user who is to be displayed enlarged together with the first user.