Nvidia Corporation

Andrew James Woodard

Benjemin Thomas Waine

Amy L Rose

Christopher Schneider

Daniel Cory Rohrer

William Bartig

Johannes Zimmerman

In various examples, as a user is speaking or presenting content during an online video conference, the data stream may be processed to generate a textual representation (e.g., transcript) of the audio and/or information relating to the video. The textual representation and/or video related information may then be processed to determine a context or one or more topic(s) of discussion. Based on the determined context/topic(s), a corresponding neural network(s) may be selected. Once a neural network has been selected, comments may be retrieved from a chat feature of the application and applied to the neural network. The neural network may then output data to indicate the relevance of the comments to the determined discussion topic. Based on the relevance of the comment, the comment may be allowed, prioritized, deleted, de-emphasized, or otherwise filtered in the chat feature.

Real-time video conference chat filtering using machine learning models

Apparatuses, systems, and techniques to detect abnormal behavior on clients using one or more neural networks on said clients. In at least one embodiment, use of or behavior on one or more clients is analyzed by a first neural network to detect abnormal behavior compared to a baseline of accepted behavior, and said baseline of accepted behavior is revised over time by a second neural network based on behavior observed on said one or more clients.

Behavior modeling using client-hosted neural networks

Apparatuses, systems, and techniques to accelerate training of one or more neural networks. In at least one embodiment, training values for individual nodes in a first neural network are estimated using a second neural network or other estimation technique in order to decrease time required to train the first neural network.

Accelerated training for neural network models

Statistical analysis can be used to attempt to identify potentially malicious references, such as trap URLs. When a URL is utilized for a request, that request can be intercepted before analysis before that URL is resolved to an address. Portions of this URL, as well as the entire URL, can be compared against one or more lists of known URLs using a probabilistic matching process to determine whether there are any matches that are very close but not quite exact. Any determined match with high probability above a suspicion threshold can be flagged as being suspicious, or associated with a potentially malicious site. An action can then be taken, such as to block that URL or prompt a user for confirmation of intent.

Detecting malformed resource references

Systems and methods are presented for reliable transmission of time-sensitive data. In particular, various embodiments provide for the generation of compressed sequential data, where individual instances of a sequence represent differentials from prior instances in that sequence. In order to reduce an amount of data that needs to be transmitted, instances of data (such as individual video frames) can be provided using a prior video frame as a reference, sending only data for those pixel locations where the pixel value differs from the reference frame. A reference frame can include a previously-received and successfully-decoded frame, in order to minimize the impact of dropped, incomplete, or corrupted frames. In order to further reduce data transmission requirements, a reference frame can be selected which is determined to be optimal for the current frame, such as may represent a least amount of data to be transmitted for a given frame.

Video compression techniques for reliable transmission

In various examples, communications having insecure protocols are dynamically hardened. For example, communications that are formatted in an outdated or otherwise insecure version of a protocol (e.g., sent by a device aged out of a service window) may be isolated within a network, converted to an updated protocol format, or any combination thereof. These systems and methods may be implemented on a general purpose network device (e.g., a hub of a Local Area Network (LAN)).

Dynamically hardening communications having insecure protocols

Annotations can be correlated to components of a document image so that the annotations track movements of the document image. A transparent overlay is generated to include the annotations and is linked to various components of the document image. Movement of the underlying document image is tracked and then adjustments in components positions are applied to associated annotations to maintain a contextual relationship between the annotations and the components of the document image.

Context aware annotations for collaborative applications

In various examples, firewalls may include machine learning models that are automatically trained and applied to analyze service inputs submitted to input processing services and to identify whether service inputs are desirable (e.g., will result in an undesirable status code if processed by a service). When a service input is determined by a firewall to be desirable, the firewall may push the service input through to the input processing service for normal processing. When a service input is determined by the firewall to be undesirable, the firewall may block or drop the service input before it reaches the input processing service and/or server. This may be used to prevent the service input, which is likely to be undesirable, from touching a server that hosts the input processing service (e.g., preventing a crash).

Application firewalls based on self-modeling service flows

Systems and methods for generating real-time synthetic crowd responses for events, to augment the experience of event participants, remote viewers, and the like. Various sensors monitor the event in question, and various event properties are derived from their output using an event state model. These event properties, along with various event parameters such as score, time remaining, etc., are then input to a machine learning model that determines a real-time synthetic audience reaction tailored to the immediate state of the event. Reaction parameters are used to generate a corresponding crowd or audience audio signal, which may be broadcast to event participants, viewers, spectators, or anyone who may be interested. This instantaneous, realistic crowd reaction more closely simulates the experience of events with full on-site audiences, enhancing the viewing experience of both event participants and those watching.

Machine learning based generation of synthetic crowd responses

Verified deliveries are commonplace for various exchanges of goods, packages, and/or other items, but often require close proximity or contact between the exchanging parties or devices associated therewith&#x2014;e.g., for digital or physical signature. To remedy this, system and methods described herein may leverage an ad hoc network established between a device of a provider and a device of a consumer for exchanging codes or tokens&#x2014;that may be validated by an authentication service&#x2014;to provide a verification process during an exchange between the parties. As a result, a safe distance may be maintained between the parties throughout the transaction&#x2014;thereby avoiding exchange of germs while also increasing safety and security of both parties&#x2014;and the verification process may be more reliable and secure.

Growing community of inventors

High Wycombe, United Kingdom

Andrew James Woodard