Abstract: Systems and methods for maintaining a project schedule in a dataset management system are provided. For example, a progress update comprising an indication of a status of training of a first machine learning algorithm may be received using at least one communication device from an external device involved in the training the first machine learning algorithm. A project schedule record may be accessed. An expected running time of the training of the first machine learning algorithm may be updated in the project schedule record based on the status of the training of the first machine learning algorithm. Information related to training of a second machine learning algorithm may be updated in the project schedule record based on the status of the training of the first machine learning algorithm.

Abstract: Systems and methods of blocking phishing attempts in a computer network, including receiving a list of assets of the computer network, wherein each asset is associated with at least one computer network address, generating at least one address permutation on the at least one computer network address of each asset of the computer network, wherein the generated at least one address permutation is different from the address associated with each asset of the computer network, receiving a communication request at a gateway server of the computer network, determining a destination address of the communication request, comparing the determined destination address with the at least one address permutation, and when the determined destination address is the same as at least one address permutation, blocking the communication request at the gateway server.

Abstract: System and method of detecting malicious interactions in a computer network, the method including generating, by a processor, at least one decoy segment, broadcasting, by the processor, the generated at least one decoy segment in a public database, monitoring, by the processor, communication within the computer network to identify interactions associated with the generated at least one decoy segment, determining, by the processor, at least one indicator of compromise (IOC) for the identified interactions, and blocking communication between the computer network and any computer associated with the determined at least one IOC.

Abstract: Systems and methods of blocking phishing attempts in a computer network, including receiving a list of assets of the computer network, wherein each asset is associated with at least one computer network address, generating at least one address permutation on the at least one computer network address of each asset of the computer network, wherein the generated at least one address permutation is different from the address associated with each asset of the computer network, receiving a communication request at a gateway server of the computer network, determining a destination address of the communication request, comparing the determined destination address with the at least one address permutation, and when the determined destination address is the same as at least one address permutation, blocking the communication request at the gateway server.

Abstract: Systems and methods of mitigating leakage of credentials of a user of a computer network, including monitoring at least one data source to scrape data that is compatible with credential data, applying a machine learning algorithm to the scraped data to identify at least one potential leaked credential, wherein the at least one potential leaked credential is identified using at least one neural network, authenticating the identified at least one potential leaked credential by a database of valid credentials of the computer network, and replacing credentials corresponding to the at least one leaked credential.

Abstract: This disclosure relates to improved techniques for performing human parsing functions using neural network architectures. The neural network architecture can model human objects in images using a hierarchal graph of interconnected nodes that correspond to anatomical features at various levels. Multi-level inference information can be generated for each of the nodes using separate inference processes. The multi-level inference information for each node can be combined or fused to generate final predictions for each of the nodes. Parsing results may be generated based on the final predictions.

Abstract: Systems and methods of mitigating leakage of credentials of a user of a computer network, including monitoring at least one data source to scrape data that is compatible with credential data, applying a machine learning algorithm to the scraped data to identify at least one potential leaked credential, wherein the at least one potential leaked credential is identified using at least one neural network, authenticating the identified at least one potential leaked credential by a database of valid credentials of the computer network, and replacing credentials corresponding to the at least one leaked credential.

Abstract: Systems and methods for enriching datasets while learning are provided. For example, intermediate results of training machine learning algorithms may be obtained. Additional training examples may be selected based on the intermediate results. In some cases, synthetic examples may be generated based on the intermediate results. The machine learning algorithms may be further trained using the selected additional training examples and/or the generated synthetic examples.

Abstract: This disclosure relates to improved techniques for performing computer vision functions, including common object detection and instance segmentation. The techniques described herein utilize neural network architectures to perform these functions in various types of images, such as natural images, UAV images, satellite images, and other images. The neural network architecture can include a dense location regression network that performs object localization and segmentation functions, at least in part, by generating offset information for multiple sub-regions of candidate object proposals, and utilizing this dense offset information to derive final predictions for locations of target objects. The neural network architecture also can include a discriminative region-of-interest (RoI) pooling network that performs classification of the localized objects, at least in part, by sampling various sub-regions of candidate proposals and performing adaptive weighting to obtain discriminative features.

Abstract: Systems and methods for selective use of examples are provided. For example, information associated with external devices may be obtained, weights may be assigned to examples using the obtained information, and inference models may be generated using the assignment of weights. In some examples, update information associated with the external devices may be obtained, the assignment of weights may be updated using the update information, and updated inference models may be generated using the updated assignment of weights.

Abstract: Systems, methods and non-transitory computer readable media for image processing are provided. Images may be obtained, for example by capturing the images using an image sensor. The images may be analyzed to identify scene information. An inference model may be obtained based on the scene information. The scene information may be used to determine parameters for modifying the obtained inference model. Additional images may be obtained, for example using the image sensor. The additional images may be analyzed using the inference model.

Abstract: System and method for image processing are provided. A stream of images may be obtained, for example by capturing images using an image sensor. Points in time associated with an activity may be obtained. For each point in time, the stream of images may be analyzed to identify events related to the activity and preceding the point in time. Based on the identified events, an event detection rule configured to analyze images to detect at least one event may be obtained. Image data may be obtained, and the image data may be analyzed using the event detection rule to detect events matching selected criteria in the image data.

Abstract: A system and a method of obtaining a location of a document on a computer network based on a document property. The method may include: receiving at least one basic marker and an encoding function associated with the document property; generating a search term according to the encoding function, based on the at least one basic marker; providing the search term to at least one search engine and obtaining therefrom one or more search results corresponding, where each search result may include one or more references to locations of documents on the computer network; discovering at least one document having the document property from the one or more search results and obtaining a discovered location of the document on the computer network; and performing at least one rule-based action, according to at least one document property of the discovered document.

Abstract: Systems and methods of blocking phishing attempts in a computer network, including receiving a list of assets of the computer network, wherein each asset is associated with at least one computer network address, generating at least one address permutation on the at least one computer network address of each asset of the computer network, wherein the generated at least one address permutation is different from the address associated with each asset of the computer network, receiving a communication request at a gateway server of the computer network, determining a destination address of the communication request, comparing the determined destination address with the at least one address permutation, and when the determined destination address is the same as at least one address permutation, blocking the communication request at the gateway server.

Abstract: Systems and methods for maintaining a project schedule in a dataset management system are provided. For example, a progress update comprising an indication of a status of training of a first machine learning algorithm may be received using at least one communication device from an external device involved in the training the first machine learning algorithm. A project schedule record may be accessed. An expected running time of the training of the first machine learning algorithm may be updated in the project schedule record based on the status of the training of the first machine learning algorithm. Information related to training of a second machine learning algorithm may be updated in the project schedule record based on the status of the training of the first machine learning algorithm.

Abstract: Systems and methods for causing actions in a dataset management system are provided. For example, a change to data maintained by the dataset management system may be detected, and a rule associated with the data and with an action may be obtained. Further, it may be determined if the change is a trigger according to the rule. In some examples, when the change is determined to be a trigger according to the rule, the action associated with the rule may be performed, for example with the data associated with the change. In some examples, when the change is determined not to be a trigger according to the rule, the system may forgo the performance of the action, for example for the data associated with the change.

Abstract: Systems and methods of mitigating leakage of credentials of a user of a computer network, including monitoring at least one data source to scrape data that is compatible with credential data, applying a machine learning algorithm to the scraped data to identify at least one potential leaked credential, wherein the at least one potential leaked credential is identified using at least one neural network, authenticating the identified at least one potential leaked credential by a database of valid credentials of the computer network, and replacing credentials corresponding to the at least one leaked credential.

Abstract: Systems and methods for merging annotations of datasets are provided. For example, assignments of labels to data-points may be obtained, confidence levels associated with the assignments of labels may be obtained. Further, the assignments of labels may be merged, for example based on the confidence levels. In some cases, inference models may be generated using the merged assignment of labels. In some examples, an update to the assignments of labels to data-points and/or the confidence levels may be obtained, and the merged assignment of labels may be updated.

Abstract: A system and a method of obtaining a location of a document on a computer network based on a document property. The method may include: receiving at least one basic marker and an encoding function associated with the document property; generating a search term according to the encoding function, based on the at least one basic marker; providing the search term to at least one search engine and obtaining therefrom one or more search results corresponding, where each search result may include one or more references to locations of documents on the computer network; discovering at least one document having the document property from the one or more search results and obtaining a discovered location of the document on the computer network; and performing at least one rule-based action, according to at least one document property of the discovered document.

Abstract: Systems and methods for determining a global update to inference models using update information from a plurality of external systems are provided. A first and a second update information associated with an inference model may be received. The first update information may be based on an analysis of a first plurality of training examples by a first external system, and the second update information may be based on an analysis of a second plurality of training examples accessible by a second external system. Information related to the distribution of the first plurality of training examples and the second plurality of training examples may be obtained. The first and the second update information, and the information related to the distribution of the first plurality of training examples and the second plurality of training examples may be used to determine a global update to the inference model.