Abstract: In various examples, one or more validity checks that model one or more aspects of human physiology may be applied to frames of detected human features to detect and respond to the presence of faults. Example validity checks include human feature constraints derived from the kinematics of human motion, anatomical and spatial constraints, consistency across detection modalities, and/or others. The present techniques may be utilized to validate human features detected by various computer vision tasks, such as those involving pose estimation, facial detection, gesture recognition, and/or activity monitoring, to name a few examples.

Abstract: In various examples, motifs, watermarks, and/or signature inputs are applied to a deep neural network (DNN) to detect faults in underlying hardware and/or software executing the DNN. Information corresponding to the motifs, watermarks, and/or signatures may be compared to the outputs of the DNN generated using the motifs, watermarks and/or signatures. When a the accuracy of the predictions are below a threshold, or do not correspond to the expected predictions of the DNN, the hardware and/or software may be determined to have a fault-such as a transient, an intermittent, or a permanent fault. Where a fault is determined, portions of the system that rely on the computations of the DNN may be shut down, or redundant systems may be used in place of the primary system. Where no fault is determined, the computations of the DNN may be relied upon by the system.

Abstract: An example computing device includes a storage device, a random-access memory, a read only memory, and a processor. The processor is to determine, during a booting process, whether an operating system being booted is a primary operating system, in response to a determination that the operating system being booted is different from the primary operating system, determine whether the operating system being booted satisfies a first criteria, and, in response to a determination that the operating system being booted fails to satisfy the first criteria, disable software persistence.

Abstract: In various examples, motifs, watermarks, and/or signature inputs are applied to a deep neural network (DNN) to detect faults in underlying hardware and/or software executing the DNN. Information corresponding to the motifs, watermarks, and/or signatures may be compared to the outputs of the DNN generated using the motifs, watermarks and/or signatures. When a the accuracy of the predictions are below a threshold, or do not correspond to the expected predictions of the DNN, the hardware and/or software may be determined to have a fault—such as a transient, an intermittent, or a permanent fault. Where a fault is determined, portions of the system that rely on the computations of the DNN may be shut down, or redundant systems may be used in place of the primary system. Where no fault is determined, the computations of the DNN may be relied upon by the system.

Abstract: In an example, a system includes a firmware controller to initiate a SM execution mode of the system. The firmware controller scans memory for a process pool tag. The firmware controller compares the process pool tag to a set of operating system process pool tags and detects a coherency discrepancy between the process pool tag and the set of operating system process pool tags. The firmware controller exits the SM execution mode of the system.

Abstract: In an example implementation according to aspects of the present disclosure, a method of a computing device may include determining whether a first mode or a second mode is selected. If the first mode is selected, the computing device connects a microphone to a sound processing logic. If the second mode is selected, the computing device disconnects the microphone from the sound processing logic and connecting an audio source to the sound processing logic.

Abstract: An example non-transitory computer readable storage medium comprising instructions that when executed cause a processor of an electronic device to: receive a password during a runtime of an operating system of the electronic device; generate a cryptographic key using the password; sign a Basic Input/Output System (BIOS) change request using the cryptographic key; and transmit the signed BIOS change request.

Abstract: In various examples, motifs, watermarks, and/or signature inputs are applied to a deep neural network (DNN) to detect faults in underlying hardware and/or software executing the DNN. Information corresponding to the motifs, watermarks, and/or signatures may be compared to the outputs of the DNN generated using the motifs, watermarks and/or signatures. When a the accuracy of the predictions are below a threshold, or do not correspond to the expected predictions of the DNN, the hardware and/or software may be determined to have a fault—such as a transient, an intermittent, or a permanent fault. Where a fault is determined, portions of the system that rely on the computations of the DNN may be shut down, or redundant systems may be used in place of the primary system. Where no fault is determined, the computations of the DNN may be relied upon by the system.

Abstract: In various examples, motifs, watermarks, and/or signature inputs are applied to a deep neural network (DNN) to detect faults in underlying hardware and/or software executing the DNN. Information corresponding to the motifs, watermarks, and/or signatures may be compared to the outputs of the DNN generated using the motifs, watermarks and/or signatures. When a the accuracy of the predictions are below a threshold, or do not correspond to the expected predictions of the DNN, the hardware and/or software may be determined to have a fault—such as a transient, an intermittent, or a permanent fault. Where a fault is determined, portions of the system that rely on the computations of the DNN may be shut down, or redundant systems may be used in place of the primary system. Where no fault is determined, the computations of the DNN may be relied upon by the system.

Abstract: In some examples, a device includes a processor, a core hardware logic to execute instructions to perform a task in the device, and a controller separate from the processor. The controller detects corruption of the instructions, and in response to detecting the corruption, load a recovery code to the core hardware logic to trigger recovery of the core hardware logic from the corruption of the instructions.

Abstract: An example computer system includes a memory having a computer program, a set of data containing a cryptographic first hash protected by a code, and a set of trusted security guidelines for operating an electronic device; a trusted platform module device to store a second hash associated with the computer program; a trusted application computing agent to establish that a hardware initialization sequence of the electronic device is trusted upon matching the first hash with the second hash; and a controller to operate the computer program on the electronic device according to the set of trusted security guidelines.

Abstract: In an example, a non-transitory computer-readable medium has instructions stored therein that, in response to being executed on computer circuitry, cause the computer circuitry to execute instructions to operate an application installed in the memory circuitry and to generate an iterative communication to indicate that the application is operating. The instructions further cause, in response to being executed, the computer circuitry to detect the presence of the iterative communication, and to reinstall the application in response to an interruption in the iterative communication.

Abstract: In an example implementation according to aspects of the present disclosure, a method of a computing device may include determining whether a first mode or a second mode is selected. If the first mode is selected, the computing device connects a microphone to a sound processing logic. If the second mode is selected, the computing device disconnects the microphone from the sound processing logic and connecting an audio source to the sound processing logic.

Abstract: According to examples, an apparatus may include a memory storing a firmware and a processor. The processor may receive a request to modify the firmware, in which the request may be associated with a first credential. The processor may also determine, based on the first credential, whether modification of the firmware is authorized and based on a determination that modification of the firmware is authorized, display a set of defined functionalities for the firmware that are authorized to be modified. The processor may further receive a modification to a functionality in the set of defined functionalities that are authorized to be modified and may apply the received modification to the functionality.

Abstract: An example computing device includes a storage device, a random-access memory, a read only memory, and a processor. The processor is to determine, during a booting process, whether an operating system being booted is a primary operating system, in response to a determination that the operating system being booted is different from the primary operating system, determine whether the operating system being booted satisfies a first criteria, and, in response to a determination that the operating system being booted fails to satisfy the first criteria, disable software persistence.

Abstract: Aspects of the present disclosure are directed to storage of network credentials. In an example, a computing apparatus for storage of network credentials includes a processor, a first memory to store an operating system of the computing apparatus, a second memory to store firmware for initiating the computing apparatus, a third memory, and a controller. Responsive to power-on of the computing apparatus, the operating system of the computing apparatus can access an agent application in the first memory, provision a wireless network credential using the agent application, and encrypt the wireless network credential using the agent application. The firmware stored by the second memory can, in response to receiving the encrypted wireless network credential from the first memory, store the encrypted wireless network credential in the third memory.

Abstract: In an example, a system includes a firmware controller to initiate a SM execution mode of the system. The firmware controller scans memory for a process pool tag. The firmware controller compares the process pool tag to a set of operating system process pool tags and detects a coherency discrepancy between the process pool tag and the set of operating system process pool tags. The firmware controller exits the SM execution mode of the system.

Abstract: In an example implementation according to aspects of the present disclosure, a method may include receiving by a controller a BIOS image, extracting a current manufacturing programming mode state corresponding to a computer system. A BIOS personality of the BIOS image is evaluated, wherein the BIOS personality is based on a differentiation of hardware and software functionality. Based on the current manufacturing programming state and the BIOS personality, a compatibility with the computer system is determined. Based on the compatibility, the BIOS image is written to non-volatile memory.

Abstract: In some examples, a device includes a processor, a core hardware logic to execute instructions to perform a task in the device, and a controller separate from the processor. The controller detects corruption of the instructions, and in response to detecting the corruption, load a recovery code to the core hardware logic to trigger recovery of the core hardware logic from the corruption of the instructions.

Abstract: An autonomous driving system could create or exacerbate a hazardous driving situation due to incorrect machine learning, algorithm design, sensor limitations, environmental conditions or other factors. This technology presents solutions that use machine learning to detect when the autonomous driving system is in this state e.g., erratic or reckless driving and other behavior, in order to take remedial action to prevent a hazard such as a collision.