Method and apparatus for employing specialist belief propagation networks
A method and apparatus for processing image data is provided. The method includes the steps of employing a main processing network for classifying one or more features of the image data, employing a monitor processing network for determining one or more confusing classifications of the image data, and spawning a specialist processing network to process image data associated with the one or more confusing classifications.
1. A method for processing multi-dimensional data, comprising the steps of:
a) acquiring one or more sets of multi-dimensional data by a main deep belief propagation network;
b) determining one or more features of the acquired multi-dimensional data;
c) determining a settled minimum error level associated with one or more of the one or more features of acquisition;
d) employing a monitoring belief propagation network to determine whether the settled minimum error level is greater than a predetermined threshold; and
e) autonomously training a new subnet to process the one or more determined features in accordance with the determination that the settled minimum error level associated with the one or more determined features is greater than the predetermined threshold.
2. The method of claim 1 , wherein the step of autonomously training the new subnet further comprises the steps of:
employing a main deep belief propagation network for classifying one or more features of the multi-dimensional data;
employing a monitor deep belief propagation network for determining one or more classifications of the multi-dimensional data; and
employing the new subnet to process multi-dimensional data associated with the one or more determined classifications.
3. The method of claim 2 , further comprising the steps of:
employing the monitor deep belief propagation network for determining one or more classifications of the multi-dimensional data by the new subnet; and
autonomously training a second new subnet to process multi-dimensional data associated with one or more of the one or more determined classifications of the multi-dimensional data processed by the new subnet.
4. He method of claim 3 further comprising the step of autonomously training a third new subnet to process data processed by the second new subnet.
5. The method of claim 2 , wherein image multi-dimensional data processed by the new subnet is provided to the main deep belief propagation network directly or via analysis and updates by the monitor deep belief propagation network.
6. The method of claim 2 , wherein the monitor deep belief propagation network employs a volatility index for determining one or more classifications of the multi-dimensional data.
7. The method of claim 2 , wherein the classifications comprise one or more objects, one or more salient features being used to train one or more of the main belief propagation networks and new subnet.
8. The method of claim 2 , wherein the classifications allow for performance of disparity decomposition via disparity decomposition metrics that change thresholds adaptively with disparity values.
9. The method of claim 8 , wherein the performance of disparity decomposition further comprises a step of extraction of energy nodules that are used in training the new subnet on various settings.
10. The method of claim 2 , further comprising the step of computing a volatility index associated with performance of disparity space decomposition in accordance with autonomously training one or more new subnets in accordance with one or more analysis and stability metrics.
11. The method of claim 2 , wherein a more comprehensive topology is employed comprising a plurality of subnets branching from the main deep belief propagation network to deal with various modalities.
12. The method of claim 11 , wherein one or more additional new subnets are utilized unrelated to either facial or gesture expression recognition.
13. The method of claim 2 , wherein the multi-dimensional data comprises image data.