Makoto Okuda

Method of shot determination in a robot camera cooperative shooting system

We are building a program-production system employing multiple robot cameras as a new program-production support technology. In this system, the robot cameras are automatically controlled in accordance with shooting rules that specify the relationship between changes in the program situation and the shots taken by individual cameras, but studio layout elements, such as the number of participants and the position in which flip-cards are displayed, are different for each program. For this reason, production staff must reset shooting rules for every program, and this operation is extremely burdensome in the limited preparation time available. We therefore devised a method of automatically generating shooting rules through simple information input based on analysis of the shooting methods of cameramen, and have tested the validity of this method in simulation tests. Moreover, we built a program-production system in which robot cameras are connected via a network to various sensors that we developed to detect changes in the program situation, and we evaluated the system by conducting program shooting experiments whose subject is engaged in actual TV program production.

A High-Accuracy Image Composition System Using a Mobile Robotic Camera

The use of robotic cameras in program production was originally limited to applications in which it was too difficult or time-consuming for humans to do the job, such as filming from high places outdoors, or for long continuous periods from a fixed location. Developments in the field of robotic technology, as well as the improved availability of high-speed camera platforms and precise sensors have resulted in advanced robotic cameras. Even today, cameras for visual effects 1,2 and robotic cameras capable of imaging fast-moving subjects 3 are used in broadcast productions. Meanwhile, television program producers continue to look for efficient ways of making their programs more appealing, with limited time, human resources, and budgets. The production of programs using robotic technology allows for the creation of appealing content, however, with the use special camera equipment, an engineer with expert knowledge of such a system and a camera operator experienced in its operation are needed when creating new video effects. — To address these problems and to implement a system capable of efficiently capturing a more diverse range of pictures, a mobile robotic camera with a motorized pedestal as well as a motorized camera platform has been developed. All of the motor axes (pan, tilt, zoom, focus, and movement) can be driven simultaneously and accurately; thus it is capable of moving in a similar manner to a real camera operator. 4–6

Real-Time LDCRF-Based Method for Inferring TV Viewer Interest

Context-awareness provides users with more versatile user interface environments when using digital devices. In a TV viewing environment, context-awareness techniques can be used for recommending useful information related to the viewed program. To incorporate a users preference in doing so, it is vital to infer whether the viewer is actually interested in the TV program. Therefore, we propose a method for inferring viewer interest in a TV program. In the proposed method, we regard the problem of inferring viewer interest as a sequential labeling problem and solve it by applying latent dynamic conditional random fields to data sequences generated by integrating the operational log information of user interaction and visual information of viewer behaviors.

Machine learning of shooting technique for controlling a robot camera

We propose a machine learning method of a TV cameramans shooting technique with a neural network so that a robot camera can automatically shoot like an experienced cameraman. An experiment using a simulator that we developed shows that, by using our method, the cameramans shooting technique can be quickly and easily embodied in the control system of the robot camera. Moreover, we derive guidelines for performing machine learning for shooting actual TV programs by analyzing the relationship of the learning data and the learning parameter with the performance of the learned neural network.