Abstract: A method is provided for generating a visualization for explaining a behavior of a machine learning (ML) model, the method includes inputting an image into a machine learning (ML) model for an inference operation. A first heatmap is generated for the image using a first visualization method. An area of highest attention is selected on the first heatmap based a predetermined threshold. The selected area is cropped from the image. The cropped selected area is upscaled. A second heatmap is generated for the cropped and upscaled selected area of the image. A final visualization is presented for analysis. In another embodiment, a computer program comprising instructions for executing the method is provided.

Abstract: A method is provided for generating a visualization for explaining a behavior of a machine learning (ML) model. In the method, an image is input to the ML model for an inference operation. The input image has an increased resolution compared to an image resolution the ML model was intended to receive as an input. A resolution of a plurality of resolution-independent convolutional layers of the neural network are adjusted because of the increased resolution of the input image. A resolution-independent convolutional layer of the neural network is selected. The selected resolution-independent convolutional layer is used to generate a plurality of activation maps. The plurality of activation maps is used in a visualization method to show what features of the image were important for the ML model to derive an inference conclusion. The method may be implemented in a computer program having instructions executable by a processor.

Abstract: A method and machine learning system for detecting adversarial examples is provided. A first machine learning model is trained with a first machine learning training data set having only training data samples with robust features. A second machine learning model is trained with a second machine learning training data set, the second machine learning training data set having only training data samples with non-robust features. A feature is a distinguishing element in a data sample. A robust feature is more resistant to adversarial perturbations than a non-robust feature. A data sample is provided to each of the first and second trained machine learning models during an inference operation. if the first trained machine learning model classifies the data sample with high confidence, and the second trained machine learning model classifies the data sample differently with a high confidence, then the data sample is determined to be an adversarial example.

Abstract: A method is provided for generating a visualization for explaining a behavior of a machine learning (ML) model, the method includes inputting an image into a machine learning (ML) model for an inference operation. A first heatmap is generated for the image using a first visualization method. An area of highest attention is selected on the first heatmap based a predetermined threshold. The selected area is cropped from the image. The cropped selected area is upscaled. A second heatmap is generated for the cropped and upscaled selected area of the image. A final visualization is presented for analysis. In another embodiment, a computer program comprising instructions for executing the method is provided.

Abstract: A method and machine learning system for detecting adversarial examples is provided. A first machine learning model is trained with a first machine learning training data set having only training data samples with robust features. A second machine learning model is trained with a second machine learning training data set, the second machine learning training data set having only training data samples with non-robust features. A feature is a distinguishing element in a data sample. A robust feature is more resistant to adversarial perturbations than a non-robust feature. A data sample is provided to each of the first and second trained machine learning models during an inference operation. if the first trained machine learning model classifies the data sample with high confidence, and the second trained machine learning model classifies the data sample differently with a high confidence, then the data sample is determined to be an adversarial example.

Abstract: A method is provided for establishing a communication session in a communications system. The method includes providing a handshake layer functional block in a first communication peer, and providing a data communication layer functional block separate from the handshake layer functional block in the first communication peer. Functionality of the data communication layer is not duplicated in the handshake layer. If the data communication layer is unable to process a received encrypted message; transmitting, by the data communication layer, a configuration request message to the handshake layer, and transmitting, by the handshake layer, in response to the configuration request message, a set channel state message to enable the data communication layer to process application data after a handshake phase of the protocol session is complete. Then, application data can be communicated through the data communication layer functional block of the first communication peer to a second communication peer.

Abstract: A method is provided for establishing a secure communication session in a communication system. The method includes providing a handshake layer functional block and providing a record layer functional block separate from the handshake layer functional block. Functionality of the record layer functional block is not duplicated in the handshake layer functional block. The record layer functional block of a first communication peer generates an ephemeral key pair. A public key of the ephemeral key pair is transmitted to the handshake layer functional block of a second communication peer via the handshake layer functional block of the first communication peer. A session key is generated from the public key of the second communication peer and a private key of the first communication peer. Messages communicated between the first communication peer and the second communication peer are protected using the session key.

Abstract: A method is provided for establishing a secure communication session in a communications system. The method includes providing a handshake layer functional block and providing a record layer functional block separate from the handshake layer functional block. A first ephemeral key pair is generated by the record layer functional block of a first communication peer. A public key of the first ephemeral key pair is transmitted to a second communication peer. The handshake layer functional block of the first communication peer generates a second ephemeral key pair. A public key of the second ephemeral key pair is transmitted to the second communication peer. The second communication peer generates a third ephemeral key pair. A handshake key is generated from the public key of the second communication peer and a private key of the handshake layer block of the first communication peer.

Abstract: A method is provided for establishing a communication session in a communications system. The method includes providing a handshake layer functional block in a first communication peer, and providing a data communication layer functional block separate from the handshake layer functional block in the first communication peer. Functionality of the data communication layer is not duplicated in the handshake layer. If the data communication layer is unable to process a received encrypted message; transmitting, by the data communication layer, a configuration request message to the handshake layer, and transmitting, by the handshake layer, in response to the configuration request message, a set channel state message to enable the data communication layer to process application data after a handshake phase of the protocol session is complete. Then, application data can be communicated through the data communication layer functional block of the first communication peer to a second communication peer.

Abstract: A method is provided for establishing a secure communication session in a communications system. The method includes providing a handshake layer functional block and providing a record layer functional block separate from the handshake layer functional block. A first ephemeral key pair is generated by the record layer functional block of a first communication peer. A public key of the first ephemeral key pair is transmitted to a second communication peer. The handshake layer functional block of the first communication peer generates a second ephemeral key pair. A public key of the second ephemeral key pair is transmitted to the second communication peer. The second communication peer generates a third ephemeral key pair. A handshake key is generated from the public key of the second communication peer and a private key of the handshake layer block of the first communication peer.

Abstract: A method is provided for establishing a secure communication session in a communication system. The method includes providing a handshake layer functional block and providing a record layer functional block separate from the handshake layer functional block. Functionality of the record layer functional block is not duplicated in the handshake layer functional block. The record layer functional block of a first communication peer generates an ephemeral key pair. A public key of the ephemeral key pair is transmitted to the handshake layer functional block of a second communication peer via the handshake layer functional block of the first communication peer. A session key is generated from the public key of the second communication peer and a private key of the first communication peer. Messages communicated between the first communication peer and the second communication peer are protected using the session key.

Abstract: A method is provided for generating a key pair and certificate for an IoT device. An integrated circuit (IC) is manufactured by a first entity for use in an internet of things (IoT) device. The IC is provided to a second entity for manufacturing the IoT device using the IC, the IC having a unique identifier (UID) and secret key derivation data (KDD). A secure memory is provided to a third entity by the first entity. The secure memory has secret key derivation parameters configured to enable the generation of a product specific parameter (PSP). The secure memory enables the third entity to prepare a signed public key certificate using the UID assigned to the IC of each device. The public key certificate is for verifying the authenticity of the device. In addition, the PSP enables the IC inside the device to generate a private key corresponding to a public key in the signed public key certificate.

Abstract: A method is provided for secure firmware provisioning of a device. In the method, an integrated circuit (IC) is manufactured by a first entity for use in the device. The IC is provided to a second entity for manufacturing the device using the IC. The IC has a unique identifier (UID) and secret key derivation data (KDD). A secure memory is provided to a third entity. The secure memory has a first key pair, and the secure memory is used with a firmware provisioning toolchain of the second entity. During manufacturing of the device by the second entity, the secure memory is enabled to verify the IC by verifying the UID. The secure memory stores a firmware decryption key, and is enabled to encrypt the firmware decryption key. The encrypted firmware decryption key is then provided to the IC, and the IC decrypts the encrypted firmware decryption key for use by the IC in decrypting the firmware.

Abstract: A method is provided for secure firmware provisioning of a device. In the method, an integrated circuit (IC) is manufactured by a first entity for use in the device. The IC is provided to a second entity for manufacturing the device using the IC. The IC has a unique identifier (UID) and secret key derivation data (KDD). A secure memory is provided to a third entity. The secure memory has a first key pair, and the secure memory is used with a firmware provisioning toolchain of the second entity. During manufacturing of the device by the second entity, the secure memory is enabled to verify the IC by verifying the UID. The secure memory stores a firmware decryption key, and is enabled to encrypt the firmware decryption key. The encrypted firmware decryption key is then provided to the IC, and the IC decrypts the encrypted firmware decryption key for use by the IC in decrypting the firmware.

Abstract: A method is provided for generating a key pair and certificate for an IoT device. An integrated circuit (IC) is manufactured by a first entity for use in an internet of things (IoT) device. The IC is provided to a second entity for manufacturing the IoT device using the IC, the IC having a unique identifier (UID) and secret key derivation data (KDD). A secure memory is provided to a third entity by the first entity. The secure memory has secret key derivation parameters configured to enable the generation of a product specific parameter (PSP). The secure memory enables the third entity to prepare a signed public key certificate using the UID assigned to the IC of each device. The public key certificate is for verifying the authenticity of the device. In addition, the PSP enables the IC inside the device to generate a private key corresponding to a public key in the signed public key certificate.