Solving inaccuracies associated with object detection

1 . A method that is computer implemented and is for solving inaccuracies associated with object detection, the method comprises:

automatically evaluating, by a controller, an accuracy of signatures for use in the object detection, the signatures were generated by an adaptable artificial intelligence (AI) system;

when finding an erroneous signature of the signatures, by the controller, triggering a generation of an error resolving part of the adaptable AI system for generating a correct signature instead of the erroneous signature, wherein the erroneous signature, when used for the object detection, results in an object detection error that is at least one of a false negative detection, a false positive detection, or inconsistent detection of objects; wherein the controller belongs to a cortical unit of the adaptable AI system, wherein the erroneous signature was generated by a signature generator of the cortical unit in response to a reception, by the cortical unit, of a sensed information unit; wherein the cortical unit further comprises a readout unit and a search engine;

receiving, during inference and by the cortical unit, a new sensed information unit;

generating, by the signature generator, a new signature;

detecting by the controller that the new signature equals the erroneous signature, by accessing error resolving rules associated with erroneous signatures;

applying an error resolving rule that is associated with the erroneous signature and with an other cortical unit of the error resolving part, by sending the new sensed information unit to the other cortical unit; and

generating, by the other cortical unit, another new signature that is not erroneous.

2 . The method according to claim 1 , comprising finding an additional erroneous signature of the signatures and triggering a generation of an additional error resolving part of the adaptable AI system for generating an additional correct signature instead of the additional erroneous signature, wherein the additional erroneous signature, when used for the object detection, results in an additional inconsistent detection of objects.

3 . The method according to claim 1 , comprising triggering a generation of the error resolving rule that associates the erroneous signature with the other cortical unit.

4 . The method according to claim 3 , wherein the method comprises extracting by the readout circuit readout information from a deep neural network (DNN) that was fed by the sensed information unit, and providing the readout information to the signature generator of the cortical unit.

5 . The method according to claim 1 , comprising generating the other cortical unit.

6 . The method according to claim 1 , comprising generating the error resolving rule.

7 . A non-transitory computer readable medium storing instructions that, when executed by at least one processor, cause the at least one processor to perform operations for solving inaccuracies associated with object detection, comprising:

automatically evaluating, by a controller, an accuracy of signatures for use in the object detection, the signatures were generated by an adaptable artificial intelligence (AI) system;

when finding an erroneous signature of the signatures, by the controller, triggering a generation of an error resolving part of the adaptable AI system for generating a correct signature instead of the erroneous signature, wherein the erroneous signature, when used for the object detection, results in an object detection error that is at least one of a false negative detection, a false positive detection, or inconsistent detection of objects; wherein the controller belongs to a cortical unit of the adaptable AI system, wherein the erroneous signature was generated by a signature generator of the cortical unit in response to a reception, by the cortical unit, of a sensed information unit of a given content; wherein the cortical unit further comprises a readout unit and a search engine;

receiving, during inference and by the cortical unit, a new sensed information unit;

generating, by the signature generator, a new signature;

detecting by the controller that the new signature equals the erroneous signature, by accessing error resolving rules associated with erroneous signatures;

applying an error resolving rule that is associated with the erroneous signature and with an other cortical unit of the error resolving part, by sending the new sensed information unit to the other cortical unit; and

generating, by the other cortical unit, another new signature that is not erroneous.

8 . The non-transitory computer readable medium according to claim 7 , storing instructions for finding an additional erroneous signature of the signatures and triggering a generation of an additional error resolving part of the adaptable AI system for generating an additional correct signature instead of the additional erroneous signature, wherein the additional erroneous signature, when used for the object detection, results in an additional inconsistent detection of objects.

9 . The non-transitory computer readable medium according to claim 7 , storing instructions for triggering a generation of the error resolving rule that associates the erroneous signature with the other cortical unit.

10 . The non-transitory computer readable medium according to claim 9 , storing instructions for extracting readout information by the readout circuit from a deep neural network (DNN) that was fed by the sensed information unit, and providing the readout information to the signature generator of the cortical unit.

11 . The non-transitory computer readable medium according to claim 7 , storing instructions for generating the other cortical unit.

12 . The non-transitory computer readable medium according to claim 7 , storing instructions for generating the error resolving rule.