Learning to be Different: Heterogeneity and Efficiency in Distributed Smart Camera Networks

Abstract

In this paper we study the self-organising behaviour of smart camera networks which use market-based handover of object tracking responsibilities to achieve an efficient allocation of objects to cameras. Specifically, we compare previously known homogeneous configurations, when all cameras use the same marketing strategy, with heterogeneous configurations, when each camera makes use of its own, possibly different marketing strategy. Our first contribution is to establish that such heterogeneity of marketing strategies can lead to system wide outcomes which are Pareto superior when compared to those possible in homogeneous configurations. However, since the particular configuration required to lead to Pareto efficiency in a given scenario will not be known in advance, our second contribution is to show how online learning of marketing strategies at the individual camera level can lead to high performing heterogeneous configurations from the system point of view, extending the Pareto front when compared to the homogeneous case. Our third contribution is to show that in many cases, the dynamic behaviour resulting from online learning leads to global outcomes which extend the Pareto front even when compared to static heterogeneous configurations. Our evaluation considers results obtained from an open source simulation package as well as data from a network of real cameras.

Self-Organising Zooms for Decentralised Redundancy Management in Visual Sensor Networks

“…This is again based on the difficulty of capturing the global goals of overlap/redundancy, proportion of tracked objects, and total confidence by using only local information. This is an instance of the problem of decomposing global goals into local rewards, when assuming decentralisation, as discussed in a related problem by Lewis et al [11], [12]. It is also important to note that our results are not perfectly even distributed in the solution space.…”

On Learning in Collective Self-Adaptive Systems: State of Practice and a 3D Framework

“…Hence, each agent should reason and act based on its own view and experiences. The opposite extreme, termed maximal knowledge access, is less common with only 8 studies [29], [31], [35], [62], [65], [72], [75], [76]. Here, all agents share complete knowledge with each other aiming to achieve the highest possible benefit for the collective.…”

“…An interesting outcome of our review is the following. In several studies [30], [35], [36], [39], [43]- [45], [48], [49], [53], [61], [63], [69], [70], the system is designed in a way that each individual selfish agent learns a "learning task" and is not aware of the fact that it is collaborating. In these circumstances, the emergent behaviour of the system is the result of the agents' unaware collaboration and is often a system-wide collaboration scheme achieving a global goal.…”

Multi-Camera Coordination and Control in Surveillance Systems

“…Their algorithm runs in a decentralized manner. Recently, Lewis et al [2013] have shown that the object tracking confidence improves when adopting heterogeneous marketing strategies by reducing the number of message exchanged. They have also shown that online learning of marketing strategies can achieved to find the suitable strategy for individual cameras.…”