An Artificial Perceptual Model Approach - Artificial Human Sensors: Science and Applications - page 118

the brain. A number of different sensor capabilities ensure the ability to connect
dynamic activities in the environment that correspond to the “attention” of the
artificial system.

Signal analysis :a perceptual procedure that organizes the received data into a
“structural picture” and extracts the requested information. This process has
similarities with the functionality in the thalamus and cerebral cortex.

•

Decision-making :an active/passive perception procedure that handles decision-
making in the system similarly to the motor cortex. In the case of artifi-
cial systems, the interpretation is performed by implementing sensing as data
acquisition added with the use of pattern recognition.

•

Knowledge base : contains all the knowledge and earlier experience as compressed in
an artificial database unit, by similar functions to a brain process.

•

Communication : communicates the information to the human user in an effective
and perceptive process. Also the information may be used as decision-making
by other artificial systems or stored in a knowledge base.

•

Human-computer interface : provide the added value for the individual when in-
tegrating the artificial information to human perception . The interaction with an ar-
tificial system and performed by human sensing, that may be involved in the
interface process. This will be discussed in more detail later in the application
part.

•

The perceptual model approach is structured in a rational organisation built
upon the human perception. However, the similarities are more on the organi-
sational level than trying to mimic the human functions. This can be seen as ad-
vantageous when both the biological, e.g., the human and artificial, e.g., perceptual
sensor systems, are intended to closely cooperate with each other in a sophisticated
way.

5.3.1 Artificial Perception Model
The conceptual model concept, earlier described in Section 5.3, has been modified
in Robertsson (2007), as shown in Fig. 5.10. This modified concept of a general
perception model was originally adapted to the use of an electronic tongue, i.e.,
gustatory sensor system. In Fig. 5.10 below, the presented model is modified by
including a possibility to implement the active or passive approach. Further, an
added feature quality is representing the set of the environmental picture, which
relates to the collected amount of measured data. The operational principle gained
by using a cognition level, when focussing on apriori definition of the feature set,
may be an attractive way in choosing the strategic approach of the method. Fur-
ther it is an advantage when shifting attention between objects or events of interest
in the active perception process mode. The arrows in between blocks in the figure
below are not representing the specific order of information flow, but a possible
interaction path. In the figure, the ability to learn from the process and explore fea-
tures is represented by the direct connecting arrow (dotted arrow) and specifically