COST: An Approach for Camera Selection and Multi-Object Inference Ordering in Dynamic Scenes
read more
Citations
A method of camera selection based on partially observable Markov decision process model in camera networks
Multi-analysis surveillance and dynamic distribution of computational resources: Towards extensible, robust, and efficient monitoring of environments
Application-driven merging and analysis of person trajectories for distributed smart camera networks
Method for estimating the visibility of features on surfaces of object instances in multi-object scenes and method for perception planning in multi-object scenes
Bargaining Strategies for Camera Selection in a Video Network
References
Loopy Belief Propagation for Approximate Inference: An Empirical Study
Loopy belief propagation for approximate inference: an empirical study
M 2 Tracker: A Multi-View Approach to Segmenting and Tracking People in a Cluttered Scene
A multiview approach to tracking people in crowded scenes using a planar homography constraint
Entropy-based sensor selection heuristic for target localization
Related Papers (5)
M 2 Tracker: A Multi-View Approach to Segmenting and Tracking People in a Cluttered Scene
Automated camera layout to satisfy task-specific and floor plan-specific coverage requirements
Frequently Asked Questions (9)
Q2. How is the algorithm used to estimate people's positions?
The algorithm cycles between using segmentation to estimate people’s ground plane positions and using ground plane position estimates to obtain segmentations; the process is iterated until stable.
Q3. How many occluded voxels can be added to a camera?
The number of occluded voxels that can be added due to dependencies depends on the selection of dependencies and the accuracy of the position estimate of the occluder.
Q4. What is the problem of a naive approach?
A naive approach (by considering all pairwise interactions of all parts of all people) would involve constructing a large Bayesian network with loops; however, this results in an intractable optimization problem.
Q5. What is the probability of a person’s visibility in a camera?
Let dV be a differential volume element(voxel) which might be included in part j of person i. The Occluder Region, Ωk(dV ), of a differential element dV in camera k is defined as the 3D region in which another person, l, must be present so that dV would not be visible in camera k (See Fig 3).
Q6. What is the probability of a person being seen in the confuser space?
The weight cl,m is proportional to the probability of the part (l,m) lying in the confuser space and being visible:cl,m = 1Z ∫ Ck(dV ) P (EO k (dA))P (El,m(dV1))dV1 (6)where Z is a normalizing factor.
Q7. How do you reduce the complexity of the information theoretic approaches?
These approaches reduce the complexity by annihilating small probabilities [11] or removing weak dependencies [14] and arcs [21].
Q8. What are the goals of a multi-perspective analysis of moving people?
Typical goals of such an analysis are to recover the position, orientation or the pose of each or some subset of the people in the scene.
Q9. What is the error in estimation of person i using the stereo pair?
the error in estimating the position of person i using the stereo pair (k1, k2) is approximated by5In M2Tracker, visibility does not vary with height and hence ground plane analysis of visibility can be performed instead of 3D modelingEi(k1, k2) = (1 − f̃(θk1,k2)Sk1i Sk2i ) (11)where θk1,k2 is the angle between the viewing directions of cameras k1 and k2 on the ground plane.