An active stereo vision-based learning approach for robotic tracking, fixating and grasping control

In this paper, an active stereo vision-based learning approach is proposed for a robot to track, fixate and grasp an object in unknown environments. First, the functional mapping relationships between the joint angles of the active stereo vision system and the spatial representations of the object are derived and expressed in a three-dimensional workspace frame. Next, the self-adaptive resonance theory-based neural networks and the feedforward neural networks are used to learn the mapping relationships in a self-organized way. Then, the approach is verified by simulation using the models of an active stereo vision system which is installed in the end-effector of a robot. Finally, the simulation results confirm the effectiveness of the present approach.

Human saccadic eye movements and tracking by active foveation in log polar space

One of the possible models of the human visual system (HVS) in the computer vision literature has a high resolution fovea and exponentially decreasing resolution periphery. The high resolution fovea is used to extract necessary information in order to solve a vision task and the periphery may be used to detect motion. To obtain the desired information, the fovea is guided by the contents of the scene and other knowledge to position the fovea over areas of interest. These eye movements are called saccades and corrective saccades. A two stage process has been implemented as a mechanism for changing foveation in log polar space. Initially, the open loop stage roughly foveates on the best interest feature and then the closed loop stage is invoked to accurately iteratively converge onto the foveation point. The open loop stage developed for the foveation algorithm is applied to saccadic eye movements and a tracking system. Log polar space is preferred over Cartesian space as: (1) it simultaneously provides high resolution and a wide viewing angle; and (2) feature invariance occurs in the fovea which simplifies the foveation process.

Tracking in a space variant active vision system

Without the ability to foveate on and maintain foveation, active vision for applications such as surveillance, object recognition and object tracking are difficult to build. Although foveation in cartesian coordinates is being actively pursued by many, multi-resolution high accuracy foveation in log polar space has not been given much attention. This paper addresses the use of foveation to track a single object as well as multiple objects for a simulated space variant active vision system. Complex logarithmic mapping is chosen firstly because it provides high resolution and wide angle viewing. Secondly, the spatially variant structure of log polar space leads to an object increasing in size as it moves towards the fovea. This is important as we know which object is closer to the fovea at any instant in time.

How is active transport associated with children's and adolescents' physical activity over time?

Background : As few longitudinal studies have examined how active transport is associated with physical activity among children and adolescents over time, and how active transport tracks through childhood and adolescence, it is important to understand whether physically active children retain their activity patterns through adolescence. This study aimed to examine (a) tracking of active transport and of moderate-to-vigorous physical activity (MVPA) across childhood and adolescence in two age cohorts; and (b) associations between active transport and MVPA at three distinct time-points, over five years.

Methods : This longitudinal study of two cohorts aged 5-6 years (n = 134) and 10-12 years (n = 201) at baseline (T1), in Melbourne, Australia, gathered follow-up data at three (T2) and five years (T3). Walking/cycling to local destinations was survey-reported; while MVPA was recorded using accelerometers and mean time spent daily in MVPA on week days and on weekends was computed. Tracking of these behaviours was examined over five years using General Estimating Equations. Linear regression analyses were performed to examine associations between active transport and MVPA at each time-point.

Results : Active transport tracked moderately among children (boys, bs = 0.36; girls, bs = 0.51) but not among adolescents. Physical activity tracked moderately (bs value range: 0.33-0.55) for both cohorts. Active transport was not associated with children’s MVPA at any time-point, but was associated with adolescent boys’ MVPA on week days at T1 (B = 1.37 (95% CI: 0.15, 2.59)), at T2 (B = 1.27 (95% CI: 0.03, 2.51)) and at T3 (B = 0.74 (95% CI: 0.01, 1.47)), and with adolescent girls’ MVPA on week days (B = 0.40 (95% CI: 0.04, 0.76)) and on weekends (B = 0.54 (95% CI:0.16, 0.93)) at T3 only.

Conclusion : Active transport was associated only with boys’ MVPA during early adolescence and with boys’ and girls’ MVPA during late adolescence. While active transport should be encouraged among all school-aged children, it may provide an important source of habitual physical activity for adolescent girls, in particular, among whom low and declining physical activity levels have been reported world-wide.

Following targets for mobile tracking in wireless sensor networks

Traditional tracking solutions in wireless sensor networks based on fixed sensors have several critical problems. First, due to the mobility of targets, a lot of sensors have to keep being active to track targets in all potential directions, which causes excessive energy consumption. Second, when there are holes in the deployment area, targets may fail to be detected when moving into holes. Third, when targets stay at certain positions for a long time, sensors surrounding them have to suffer heavier work pressure than do others, which leads to a bottleneck for the entire network. To solve these problems, a few mobile sensors are introduced to follow targets directly for tracking because the energy capacity of mobile sensors is less constrained and they can detect targets closely with high tracking quality. Based on a realistic detection model, a solution of scheduling mobile sensors and fixed sensors for target tracking is proposed. Moreover, the movement path of mobile sensors has a provable performance bound compared to the optimal solution. Results of extensive simulations show that mobile sensors can improve tracking quality even if holes exist in the area and can reduce energy consumption of sensors effectively.