Review of techniques for motion capture data processing

In order to high reality and efficiency, the technique computer animation. With the development of motion capture, a of motion capture （MoCap） has been widely used in the field of large amount of motion capture databases are available and this is significant for the reuse of motion data. But due to the high degree of freedoms and high capture frequency, the dimension of the mo- tion capture data is usually very high and this will lead to a low efficiency in data processing. So how to process the high dimension data and design an efficient and effective retrieval approach has become a challenge which we can＇t ignore. In this paper, first we lay out some problems about the key techniques in motion capture data processing. Then the existing approaches are analyzed and sum- marized. At last, some future work is proposed.

Key-frame reference selection for non-feedback video communication

This article addresses the problem of reference picture optimization in video communication over error prone networks. A novel estimation model for transmission distortion is proposed. This model is capable of recursively estimating the overall end-to-end distortion caused by quantization, error propagation, and error concealment. Simulation results show that this model can accurately estimate channel distortion. Then, based on the distortion estimation model, a new non-feedback key-frame reference picture selection （KRPS） algorithm is developed. The optimum reference picture minimizes the transmission distortion under the rate-distortion optimization framework. Extensive experiment results demonstrate that the proposed KRPS algorithm substantially achieves more peak signal to noise ratio （PSNR） gain over traditional prediction, especially in low bit-rate transmission.

Image label transfer: Short video labelling by using frame auto-encoder

Short videos on the Internet have a huge amount, but most of them are unlabeled. In this paper, a rough short video labelling method based on the image classification neural network is proposed. Convolutional auto-encoder is applied to train and learn unlabeled video frames, in order to obtain feature in the specific level. With these features, the video key-frames are extracted by the feature clustering method. These key-frames which represent the video content are put into an image classification network, so that the labels of every video clip can be got. In addition, the different architectures of convolutional auto-encoder are estimated, and a better performance architecture through the experiment result is selected. In the final experiment, the video frame features from the convolutional auto-encoder are compared with those from other extraction methods, where it illustrates remarkable results by the proposed method.

Movie Scene Recognition Using Panoramic Frame and Representative Feature Patches

Recognizing scene information in images or has attracted much attention in computer vision or videos, such as locating the objects and answering ＂Where am research field. Many existing scene recognition methods focus on static images, and cannot achieve satisfactory results on videos which contain more complex scenes features than images. In this paper, we propose a robust movie scene recognition approach based on panoramic frame and representative feature patch. More specifically, the movie is first efficiently segmented into video shots and scenes. Secondly, we introduce a novel key-frame extraction method using panoramic frame and also a local feature extraction process is applied to get the representative feature patches （RFPs） in each video shot. Thirdly, a Latent Dirichlet Allocation （LDA） based recognition model is trained to recognize the scene within each individual video scene clip. The correlations between video clips are considered to enhance the recognition performance. When our proposed approach is implemented to recognize the scene in realistic movies, the experimental results shows that it can achieve satisfactory performance.