Abstract: A certification method, system, and computer program product include certifying an adversarial robustness of a convolutional neural network by deriving an analytic solution for a neural network output using an efficient upper bound and an efficient lower bound on an activation function and applying the analytic solution in computing a certified robustness.

Abstract: The present disclosure relates to a magnetic liquid sealing device. When a rotating shaft of the device axially vibrates, an outer ring of a bearing may not be provided with a rim, and a J-shaped bush and U-shaped sleeve ring can have relative displacement in an axial direction. When the rotating shaft radially vibrates, the U-shaped sleeve ring may deviate radially to allow sealing gaps, so as to prevent a bump between a pole shoe and J-shaped bush. Under action of a support spring and axial spring, sealing rings can be pressed to allow for sealing, and the U-shaped sleeve ring may not be in direct contact with an end cap, so as to avoid face friction.

Abstract: A power battery heating method for an electric vehicle includes: acquiring a heating power demand of a power battery; acquiring power demand information of a driving module of the electric vehicle in real time, and determining a current heating power of the power battery according to the power demand information; acquiring a compensating heating current according to the heating power demand and the current heating power when the current heating power is less than the heating power demand; causing the motor controller to regulate a control current of the driving motor according to the compensating heating current, so that the driving motor outputs a high-frequency oscillation current equal to the compensating heating current; and causing the power battery to perform self-heating according to the high-frequency oscillation current outputted by the driving motor.

Abstract: A warning method for battery thermal runaway includes: acquiring current temperature data of each temperature sensor arranged in a battery pack, current voltage data of each voltage sensor arranged in the battery pack, and current sensor disconnection data of each temperature sensor; calculating a number of occurrences of a primary feature at a current moment according to the current temperature data; calculating a number of occurrences of a secondary feature at the current moment according to the current voltage data and the current sensor disconnection data in a case that the number of occurrences of the primary feature is greater than 0; calculating a sum of the number of occurrences of the primary feature and the number of occurrences of the secondary feature at the current moment; and sending a warning for battery thermal runaway when the sum is greater than a preset feature number threshold.

Abstract: The present disclosure relates to a magnetic liquid sealing device. When a rotating shaft of the device axially vibrates, an outer ring of a bearing may not be provided with a rim, and a J-shaped bush and U-shaped sleeve ring can have relative displacement in an axial direction. When the rotating shaft radially vibrates, the U-shaped sleeve ring may deviate radially to allow sealing gaps, so as to prevent a bump between a pole shoe and J-shaped bush. Under action of a support spring and axial spring, sealing rings can be pressed to allow for sealing, and the U-shaped sleeve ring may not be in direct contact with an end cap, so as to avoid face friction.

Abstract: A battery internal temperature information processing method, a computer device, and a storage medium that first acquire off-line testing data for off-line testing a battery module and construct an equivalent thermal network model from the off-line testing data, determine optimal model parameters of the equivalent thermal network model based on a multi-objective function fitting method; thereafter, a first battery internal temperature estimate of the battery of the vehicle at a first moment in actual operation of the vehicle is determined, in turn, based on the acquired initial state vector values of the battery of the vehicle, first operational data at a first moment in actual operation of the vehicle, and an equivalent thermal network model including the optimal model parameters.

Abstract: Adversarial patches can be inserted into sample pictures by an adversarial image generator to realistically depict adversarial images. The adversarial image generator can be utilized to train an adversarial patch generator by inserting generated patches into sample pictures, and submitting the resulting adversarial images to object detection models. This way, the adversarial patch generator can be trained to generate patches capable of defeating object detection models.

Abstract: A hybrid-distance adversarial patch generator can be trained to generate a hybrid adversarial patch effective at multiple distances. The hybrid patch can be inserted into multiple sample images, each depicting an object, to simulate inclusion of the hybrid patch at multiple distances. The multiple sample images can then be used to train an object detection model to detect the objects.

Abstract: A graph neural network can be built and trained to predict a risk of an entity. A multi-relational graph network can include a first graph network and a second graph network. The first graph network can include a first set of nodes and a first set of edges connecting some of the nodes in the first set. The second graph network can include a second set of nodes and a second set of edges connecting some of the nodes in the second set. The first set of nodes and the second set of nodes can represent entities, the first set of edges can represent a first relationship between the entities and the second set of edges can represent a second relationship between the entities. A graph convolutional network (GCN) can be structured to incorporate the multi-relational graph network, and trained to predict a risk associated with a given entity.

Abstract: A battery temperature estimation method includes fitting, when a battery is in an offline state, a first function relationship according to corresponding admittances of the battery at different test temperatures, and obtaining a temperature distribution model of the battery according to the shape and the size of the battery, and determining a second function relationship corresponding to internal temperatures and a surface temperature by combining with the first function relationship. The second function relationship is used for estimating the internal temperature of the battery by using the surface temperature and the admittances of the battery.

Abstract: A step-varying equalization method, a device, a medium, a battery pack, and a vehicle are provided. The method includes: initiating coarse-tuning equalization for a cell in the series-connected battery when an initial equalization difference of the cell reaches a preset coarse-tuning requirement; determining a first state of charge (SOC) equalization difference according to a first voltage value of the cell after completion of the coarse-tuning equalization when a first real equalization difference of the cell after the coarse-tuning equalization reaches a preset fine-tuning requirement, and initiating fine-tuning equalization for the cell with a first equalization step size based on the first SOC equalization difference; and determining that SOC equalization of the cell is completed when a second real equalization difference of the cell after completion of the fine-tuning equalization is less than or equal to a target equalization value.

Abstract: Method and apparatus that includes receiving a query describing an aspect in a video, the video including a plurality of frames, identifying multiple proposals that potentially correspond to the query where each of the proposals includes a subset of the plurality of frames, ranking the proposals using a graph convolution network that identifies relationships between the proposals, and selecting, based on the ranking, one of the proposals as a video segment that correlates to the query.

Abstract: An illustrative embodiment includes a method for protecting a machine learning model. The method includes: determining concept-level interpretability of respective units within the model; determining sensitivity of the respective units within the model to an adversarial attack; identifying units within the model which are both interpretable and sensitive to the adversarial attack; and enhancing defense against the adversarial attack by masking at least a portion of the units identified as both interpretable and sensitive to the adversarial attack.

Abstract: Scalable distributed adversarial training techniques for robust deep neural networks are provided. In one aspect, a method for adversarial training of a deep neural network-based model by distributed computing machines M includes, by distributed computing machines M: obtaining adversarial perturbation-modified training examples for samples in a local dataset D(i); computing gradients of a local cost function fi with respect to parameters ? of the deep neural network-based model using the adversarial perturbation-modified training examples; transmitting the gradients of the local cost function fi to a server which aggregates the gradients of the local cost function fi and transmits an aggregated gradient to the distributed computing machines M; and updating the parameters ? of the deep neural network-based model stored at each of the distributed computing machines M based on the aggregated gradient received from the server.

Abstract: Techniques for training robust machine learning models for adversarial input data. Training data for a machine learning (ML) model is received. The training data includes a plurality of labels for data elements. First modified training data is generated by modifying one or more of the plurality of labels in the training data using parameterized label smoothing with a first optimization parameter. The ML model is trained using the first modified training data.

Abstract: A trained machine learning model and a training dataset used to train the trained machine learning model can be received. Based on the training dataset, unsupervised adversarial examples can be generated. Robustness of the trained machine learning model can be determined using the generated unsupervised adversarial examples. The training dataset can be augmented with the generated unsupervised adversarial examples. The trained machine learning model can be retrained using the augmented training dataset.

Abstract: Embodiments relate to a system, program product, and method to support a convolutional neural network (CNN). A class-specific discriminative image region is localized to interpret a prediction of a CNN and to apply a class activation map (CAM) function to received input data. First and second attacks are generated on the CNN with respect to the received input data. The first attack generates first perturbed data and a corresponding first CAM, and the second attack generates second perturbed data and a corresponding second CAM. An interpretability discrepancy is measured to quantify one or more differences between the first CAM and the second CAM. The measured interpretability discrepancy is applied to the CNN. The application is a response to an inconsistency between the first CAM and the second CAM and functions to strengthen the CNN against an adversarial attack.

Abstract: One or more computer processors generate a plurality of adversarial perturbations associated with a model, wherein the plurality of adversarial perturbations comprises a universal perturbation and one or more per-sample perturbations. The one or more computer processors identify a plurality of neuron activations associated with the model and the plurality of generated adversarial perturbations. The one or more computer processors maximize the identified plurality of neuron activations. The one or more computer processors determine the model is a Trojan model by leveraging one or more similarities associated with the maximized neuron activations and the generated adversarial perturbations.

Abstract: Systems and methods for performing video understanding and analysis. Sets of feature maps for high resolution images and low resolution images in a time sequence of images are combined into combined sets of feature maps each having N feature maps. A time sequence of temporally aggregated sets of feature maps is created for each combined set of feature maps by: selecting a selected combined set of feature maps corresponding to an image at time “t” in the time sequence of images; applying, by channel-wise multiplication, a feature map weighting vector to a number of combined sets of feature maps that are temporally adjacent to the selected combined set of feature maps; and summing elements of the number of combined set of feature maps into a temporally aggregated set of feature maps. The time sequence of temporally aggregated sets of feature maps is processed to perform video understanding processing.

Abstract: Interpretation maps of deep neural networks are provided that use Renyi differential privacy to guarantee the robustness of the interpretation. In one aspect, a method for generating interpretation maps with guaranteed robustness includes: perturbing an original digital image by adding Gaussian noise to the original digital image to obtain m noisy images; providing the m noisy images as input to a deep neural network; interpreting output from the deep neural network to obtain m noisy interpretations corresponding to the m noisy images; thresholding the m noisy interpretations to obtain a top-k of the m noisy interpretations; and averaging the top-k of the m noisy interpretations to produce an interpretation map with certifiable robustness.