Hi, I am Larry. Welcome to my blog   :-)
I am a research scientist focusing on data analytics on mobile devices, especially topics of real-time computer vision and multi-modality signal processing for mobile sensors. This blog is created to share my research experience, to explain some technical issues I have solved, and to bring you some useful information. A blog is also a platform to share knowledge to audiences around the world, so your feedback is always welcome.
An image tracking device and an image tracking method thereof are provided. The image tracking device includes an image capture interface, a storage means, and a processor means. The storage means has a multi-dimensional storage space for storing a plurality of first images, each dimension of the multi-dimensional storage space being corresponding to a feature-related variance of a multi-dimensional variance. The processor means is configured to execute the following operations: marking a second image in the picture frame; calculating a multi-dimensional variance between the second image and each of the first images separately; determining whether the second image contains the object according to the multi-dimensional variance calculated; and if the second image is determined as one containing the object, storing the second image as one of the first images, in a specific subspace of the multi-dimensional storage space according to the multi-dimensional variance calculated.
The present invention is related to a social system and process used for bringing virtual social network into real life, which is allowed for gathering and analyzing a social message of at least one interlocutor from virtual social network so as to generate at least one recommended topic, allowing a user to talk with the interlocutor through the utilization of the recommended topic, and then capturing and analyzing at least one speech and behavior and/or physiological response of the interlocutor during talking so as to generate an emotion state of the interlocutor. The user is allowed to determine whether the interlocutor is interested in the recommended topic through the emotion state of interlocutor. Thus, it is possible to bring the social message on virtual network into real life, so as to increase communication topics between persons in real life.
A method for scaling video content based on bandwidth rate is disclosed. Wherein, a plurality of target shots and non-target shots are recognized from shots of a video. A background and at least one foreground object are separated from the video frame of the target shot. The present invention method provides a plurality of bandwidth levels, and one among them is selected in advance. At least one foreground object is selected and pasted on the background based on the selected bandwidth level, so as to obtain at least one recombined target shot. The recombined target shot, the non-target shot or other recombined target shot can be merged to form a recombined video unit based on the selected bandwidth level. Therefore, the present invention proposes video content of scalability based on bandwidth rate to provide the different video bit-stream size, so that viewer can get the high quality video under the lower bandwidth rate.