Apparatus and method for monitoring and controlling of a neural network using another neural network implemented on one or more solid-state chips
A method of operating an apparatus using a control system that includes at least one neural network. The method includes receiving an input value captured by the apparatus, processing the input value using the at least one neural network of the control system implemented on first one or more solid-state chips, and obtaining an output from the at least one neural network resulting from processing the input value. The method may also include processing the output with another neural network implemented on solid-state chips to determine whether the output breaches a predetermined condition that is unchangeable after an initial installation onto the control system. The aforementioned another neural network is prevented from being retrained. The method may also include the step of using the output from the at least one neural network to control the apparatus unless the output breaches the predetermined condition. Similar corresponding apparatuses are described.
1. A method of operating an apparatus using a control system, the method comprising:
receiving an input value from an apparatus, the input value based on an image captured by a device coupled to the apparatus;
processing the input value using at least one neural network of the control system, the at least one neural network implemented on a first one or more solid-state chips, and obtaining an output from the at least one neural network resulting from processing the input value;
processing the output with a second neural network of the control system implemented on a second one or more solid-state chips to determine whether the output breaches a predetermined condition that is unchangeable after an initial installation onto the control system, wherein the second neural network is prevented from being retrained;
re-training the at least one neural network with a second training data set of images that is different than a first training set of images used to initially train the at least one neural network, when the output is determined to breach the predetermined condition; and
using the output from the at least one neural network to control the apparatus unless the output breaches the predetermined condition.
2. The method of claim 1 , wherein determining whether the obtained output breaches the predetermined condition comprises determining whether the obtained output breaches one of a set of predetermined conditions, the set of predetermined conditions including a first boundary condition and a second boundary condition, wherein in response to determining the obtained output breaches the first boundary condition, ignoring the obtained output and not using the output to control the apparatus, and wherein in response to breaching the second boundary condition, shutting down the apparatus or providing a notification that the machine needs to be used in a manual mode to prevent a damage to the apparatus or someone, or something, in the area of the apparatus.
3. The method of claim 1 , wherein the apparatus is an autonomous land vehicle and the step of using the obtained output further includes the step of generating a signal to control the autonomous land vehicle.
4. The method of claim 1 , further comprising:
in response to determining the obtained output from the at least one neural network breaches a predetermined condition, replacing nodal values of the at least one neural network with previously stored nodal values.
5. The apparatus of claim 1 , wherein the at least one neural network is trained with a first data set and the second neural network is trained with a second data set, the second data set different than the first data set.
6. An apparatus operated in part by a controller, the apparatus comprising:
an input device coupled to the apparatus, the input device constructed to generate an input value based on an image captured by the input device;
a controller;
at least one neural network implemented on a first one or more solid-state chips coupled to the controller, the at least one neural network constructed to receive the input value and to generate an output; and
wherein the controller comprises a second neural network implemented on a second one or more solid-state chips constructed to receive the output from the at least one neural network and determine whether the output breaches a predetermined condition unchangeable after an initial installation onto the controller,
wherein the second neural network is prevented from being retrained,
wherein the controller is constructed to operate the apparatus using the output from the at least one neural network unless the output from the at least one neural network is determined to breach the predetermined condition, and
wherein the controller is further constructed to re-train the at least one neural network, with an image training data set that is different than an initial image training data set used to train the at least one neural, when the output is determined to breach the predetermined condition.
7. The apparatus of claim 6 , wherein the controller is further constructed to generate a human recognizable notification when the output is determined to breach the predetermined condition.
8. The apparatus of claim 6 , wherein the controller is further constructed to determine whether the output breaches one of a set of predetermined conditions, the set of predetermined conditions including a first boundary condition and a second boundary condition, wherein in response to determining the output breaches the first boundary condition the output is not used to control the apparatus, and wherein in response to determining the output breaches the second boundary condition, the controller shuts down the apparatus, or provides a notification that the machine needs to be used in a manual mode, to prevent damage to the apparatus or someone or something in the area of the apparatus.
9. The apparatus of claim 6 , wherein the apparatus is an autonomous land vehicle coupled to the at least one neural network and the controller is constructed to generate a signal based on the output to control the autonomous land vehicle.
10. The apparatus of claim 6 , wherein nodal values of the at least one neural network are replaced by previously stored nodal values when the obtained output from the at least one neural network is determined to breach a predetermined condition.
11. The apparatus of claim 6 , wherein the at least one neural network is trained with a first data set and the second neural network is trained with a second data set, the second data set different than the first data set.