Much of the research on face recognition has focused so far only on the identification from full frontal/profile facial images. As partial faces, due to occlusion, are all what is sometimes available for the Face in a...
详细信息
Much of the research on face recognition has focused so far only on the identification from full frontal/profile facial images. As partial faces, due to occlusion, are all what is sometimes available for the Face in a Crowd challenge, we construct composite faces by augmenting either the left or right half of the face, with its mirror image. Testing has to match then the composite faces available against the full face and vice versa. Training is done using a mixture of experts, the Ensemble of Radial Basis Functions (ERBF), which can cope with the inherent variability in the image formation and data acquisition process. The database used for experiments consists of 150 different subjects whose pose ranges across ± 5° rotation to yield a total of 3,000 face images. Our experimental results, using k-fold Cross Validation (CV), show the feasibility and usefulness of our method for the Face in a Crowd task. The best recognition rates, obtained when either the composite left or right face is matched against the full face, yield 94% and 90%, while we obtain 96% and 92% when the full face is matched against either the composite left or right face.
A differential electromagnetic transducer (DET), which has similar frequency characteristics to those of the normal middle ear, has been designed and implemented for an implantable middle ear (IME). In order to optimi...
详细信息
This paper deals with understanding referring expressions involving perceptual grouping. The ability to use referring expressions is important for conversational agents aimed at real-world interaction. We conducted a ...
详细信息
This paper deals with understanding referring expressions involving perceptual grouping. The ability to use referring expressions is important for conversational agents aimed at real-world interaction. We conducted a psychological experiment to collect referring expressions involving perceptual grouping. A set of methods to identify referents based on the collected data are presented. We were able to identify 78.8% of the referents in the collected expressions
This paper examines MPEG-4 coding efficiency for interlaced and progressively scanned video data recorded during remote patient monitoring. Experiments are performed using both interlaced and progressive coding method...
详细信息
This paper examines MPEG-4 coding efficiency for interlaced and progressively scanned video data recorded during remote patient monitoring. Experiments are performed using both interlaced and progressive coding methods to compress interlaced, deinterlaced and progressively scanned video sequences at different bit rates. Our results show that progressively scanned video outperforms interlaced video in both coding efficiency and video quality. We also found that the use of a deinterlacing algorithm improves video coding performance. In addition, we found that interlaced or progressively scanned video data can be compressed using either progressive or interlaced coding algorithms without significant difference in performance.
Advanced driver assistance systems are designed to make driving easier that is, to alleviate the driver's workload, and to increase traffic safety. However, traffic safety is affected by negative behavioral adapta...
详细信息
Advanced driver assistance systems are designed to make driving easier that is, to alleviate the driver's workload, and to increase traffic safety. However, traffic safety is affected by negative behavioral adaptation, meaning that drivers tend to increase speed and pay less attention to driving when supported by an advanced assistance system. We relate behavioral adaptation to reinforcement learning at a subconscious level, and propose that driver assistance is dynamically varied within predetermined safety limits. The aim of employing a dynamic assistance policy is to prevent the driver from noticing a constant improvement in vehicle handling. We conclude by describing ongoing work for empirically evaluating an improved lane departure warning system that uses a dynamic assistance policy.
A humanoid robot is a real-time system controlled by a complex computer system that requires huge computing power for perception and planning, high energy efficiency for self-contained control, reduction of physical d...
详细信息
A humanoid robot is a real-time system controlled by a complex computer system that requires huge computing power for perception and planning, high energy efficiency for self-contained control, reduction of physical dimensions, and high reliability. This paper proposes a distributed architecture for the humanoid robot control substituting conventional centralized control architectures. In addition to the parallelism that provides scalable computing power at low clock namely at low energy, the distributed architecture contributes to reliable operations by replacing many fragile analog signal wires with a digital network with redundant routes. In order to accomplish a real-time control over the network, RMTP (responsive multi-threaded processor) for parallel and real-time computation has been newly designed. RMTP can synchronize more than thirty nodes distributed over a robot body in less than 5 micro second with a real time network called the responsive link (RL). Architectures of RMTP, RL and Linux-based real-time system software are presented.
This paper presents a neural network ensemble creation method where component networks are determined automatically by sequential training. Only previously misclassified patterns by existing component networks are use...
详细信息
This paper presents a neural network ensemble creation method where component networks are determined automatically by sequential training. Only previously misclassified patterns by existing component networks are used for training coming component networks sequentially. This training strategy forces coming component networks to work only on the unsolved portion of the input space. As a result coming component networks could easily maintain diversity with existing component networks in the ensemble. Finally all component networks are trained simultaneously, while maintaining error interaction among them through a penalty function. This new method has been tested extensively on several benchmark problems of machine learning and neural networks. Experimental result shows that it can produce compact ensemble structure that exhibits good generalization ability
The developing of an interactive augmented reality (AR) application with the help of various authoring tool is discussed. Building an interactive AR application involves two main steps, creating graphics and audio con...
详细信息
The developing of an interactive augmented reality (AR) application with the help of various authoring tool is discussed. Building an interactive AR application involves two main steps, creating graphics and audio content and describing object behaviors and the interactions between users and that content. The Designers Augmented Reality Toolkit (DART) is a tool for the rapid prototyping of AR applications and provides event-trigger components. Immersive authoring with 'What You eXperience is What You Get (WYXIWUG) allows users to experience and verify immersive content firsthand, creating it through natural and direct interaction within the same environment.
暂无评论