Multisensor Fusion and Integration seminars report

[seminar class]

[attachment=11540]
Architecture
(US-JDL/UK-TFDF)
Feature Space
(Data representations, Task-specific, feedback)
Dimensionality
(Communication bandwidth constraints,High [ Low, increase SnR
Uncertainty
 Data, sensor, communication noise, high level ignorance, model uncertainty
 `soft’ decisions – Bayesian inference framework… but ….
 Incorrect use of independence between models [ Veto Effect
 Inaccurate estimation of probabilities can lead to severe distortion of decisions (product rule dominated by low probability errors)
 Simpler decision methods more robust
Dynamics
 Fusion is an iterative dynamical process - Continually refining estimates, representations ..
 How do constraints on communication bandwidth and processing limit architectures for fusion?
 How does the brain create and modify its data representation?
 How does the brain encode time, dynamics and use feedback?
 How does the brain encode and process probabilities and uncertain knowledge?
 Apart from very low level (cellular/subcellular) and very high level binding, the brain appears to leave data sources fragmented. Why?
 (interesting clinical exception in synaesthesia! – do we learn ICA?)