DP-FACT: Towards topological mapping and scene recognition with color for omnidirectional camera

Abstract

Abstract-Topological mapping and scene recognition problems are still challenging, especially for online realtime visionbased applications. We develop a hierarchical probabilistic model to tackle them using color information. This work is stimulated by our previous work [1] which defined a lightweight descriptor using color and geometry information from segmented panoramic images. Our novel model uses a Dirichlet Process Mixture Model to combine color and geometry features which are extracted from omnidirectional images. The inference of the model is based on an approximation of conditional probabilities of observations given estimated models. It allows online inference of the mixture model in real-time (at 50Hz), which outperforms other existing approaches. A real experiment is carried out on a mobile robot equipped with an omnidirectional camera. The results show the competence against the state-of-art.

DP-Fusion: A generic framework for online multi sensor recognition

“…Experiment results on a real-time scene recognition problem, by fusing two different types of readings, can be obtained from our previous report [27].…”

“…For all N readings in the sequence, x p t represents the perceived information acquired at time-stamp t from sensor p. Taking discretized readings as an example, perceived information from raw sensor data can be represented as histograms [27]. Assuming there are K different clusters, θ k is then a matrix of K × Z p , where Z p is the number of possible histograms for sensor p. x t 's of dimension Z p are drawn from θ k .…”

“…We instantiated a similar model as scene recognition problem in [27] for dual-sensor perception. Based on further study, we draw the following properties of the model.…”

“…The complexity of the algorithm only rises linearly along with the number of models. Based on the dual-sensor model in [27], we record the computation time over the test. The result is shown in figure 5.…”

A Pricing Mechanism for Task Oriented Resource Allocation in Cloud Robotics

“…nition [40] and scene analysis [39,42]. However, developing a practical robot that can cover many services would be extremely expensive and require a long time.…”