Sorry, you need to enable JavaScript to visit this website.

Rich image and text features can largely improve the training of image captioning tasks. However, rich image and text features mean the incorporation of a large amount of unnecessary information. In our work, in order to fully explore and utilize the key information in images and text, we view the combination of image and text features as a data screening problem. The combination of image and text features is dynamically screened through a series of inference strategies with the aim of selecting the optimal image and text features.

Low latency video streaming of volumetric content is an emerging technology to enable immersive media experiences on mobile devices. Unlike 3DoF scenarios where users are restricted to changes of their head orientation at a single position, volumetric content allows users to move freely within the scene in 6DoF. Although the processing power of mobile devices has increased considerably, streaming volumetric content directly to such devices is still challenging. High-quality volumetric content requires significant data rate and network bandwidth.

This is an overview poster on recent research of visual analysis algorithms in visual sensor networks. With the following abstract:

Communication during emergency situations is crucial to saving lives. Rescue workers at an emergency scene need to be able to coordinate and communicate effectively. Despite the vast improvements in personal communication networks, public safety communication has been lacking. A recent bill from Congress and the FCC has provided the groundwork for the creation of a nationwide broadband public safety communication network. This advancement in technology will allow rescue workers to receive critical information updates in all forms of media (e.g., video, text and voice).

Rate control is very crucial for scalable video applications as it controls encoding bitrates so as to meet channel bandwidths while obtaining optimum encoding quality. In this paper, we proposed a new inter-layer rate control algorithm for H.264/SVC. One unique feature is that an effective switched model is proposed to predict the bit estimation used for encoding inter frames either from the previous frame of the current layer or from the current frame of the previous layer.