Systems and methods for determining a location based on image data are provided. A method can include receiving, by a computing system, a query image depicting a surrounding environment of a vehicle. The query image can be input into a machine-learned image embedding model and a machine-learned feature extraction model to obtain a query embedding and a query feature representation, respectively. The method can include identifying a subset of candidate embeddings that have embeddings similar to the query embedding. The method can include obtaining a respective feature representation for each image associated with the subset of candidate embeddings. The method can include determining a set of relative displacements between each image associated with the subset of candidate embeddings and the query image and determining a localized state of a vehicle based at least in part on the set of relative displacements.

Disclosed herein include systems, devices, computer readable media, and methods for resilience determination and damage recovery in neural networks using a weight space and a metric that together form a manifold (such as a pseudo-Riemannian manifold or a Riemannian manifold).

In one embodiment, a set of feature vectors can be derived from any biometric data, and then using a deep neural network (DNN) on those one-way homomorphic encryptions (i.e., each biometrics' feature vector) can determine matches or execute searches on encrypted data. Each biometrics' feature vector can then be stored and/or used in conjunction with respective classifications, for use in subsequent comparisons without fear of compromising the original biometric data. In various embodiments, the original biometric data is discarded responsive to generating the encrypted values. In another embodiment, the homomorphic encryption enables computations and comparisons on cypher text without decryption. This improves security over conventional approaches. Searching biometrics in the clear on any system, represents a significant security vulnerability. In various examples described herein, only the one-way encrypted biometric data is available on a given device. Various embodiments restrict execution to occur on encrypted biometrics for any matching or searching.

The operations of a computer-implemented method include obtaining a topological map of an environment including a series of waypoints and a series of edges. Each edge topologically connects a corresponding pair of adjacent waypoints. The edges represent traversable routes for a robot. The operations include determining, using the topological map and sensor data captured by the robot, one or more candidate alternate edges. Each candidate alternate edge potentially connects a corresponding pair of waypoints that are not connected by one of the edges. For each respective candidate alternate edge, the operations include determining, using the sensor data, whether the robot can traverse the respective candidate alternate edge without colliding with an obstacle and, when the robot can traverse the respective candidate alternate edge, confirming the respective candidate alternate edge as a respective alternate edge. The operations include updating, using nonlinear optimization and the confirmed alternate edges, the topological map.

Systems, methods, and computer readable media for data augmentation are described. The system comprises a network device, a memory comprising a data augmentation model and a plurality of seed entries, and a processor in communication with the network device and the memory. The processor is configured to receive a candidate data item in a second data set, generate a candidate seed corresponding to the candidate data item, and determine a data feature, based on the data augmentation model, for the candidate seed. Additionally, the processor is configured to generate at least one matching seed in the plurality of seed entries, the at least one matching seed based on the data feature. The processor is further configured to augment the candidate data item with data corresponding to the at least one matching seed.

The disclosure generally relates to a computer-implemented method for training a machine learning model for text generation, the method comprising inputting text into the machine learning model; preprocessing the input text to obtain a plurality of character vector representations; encoding, using an encoder, each of the plurality of character vector representations to obtain a plurality of word vector representations; generating, using a backbone model, a plurality of predictive word vector representations based on the plurality of word vector representations; decoding, using a decoder, the plurality of predictive word vector representations to obtain a plurality of character-probabilities; and updating the machine learning model based the plurality of character-probabilities. The disclosure also relates to a computer implemented method for generating text, a corresponding device, system and computer program.

A continual learning method includes receiving multimodality data including data items having modalities and tokenizing each data, receiving text data representing a class for the multimodality data and tokenizing the text data, generating an aggregated modality prompt, generating an aggregated text prompt, inputting modality concatenation data, in which the aggregated modality prompt is concatenated with the tokenized multimodality data, into a vision encoder and outputting a modality embedding vector, inputting text concatenation data in which the aggregated text prompt is concatenated with the tokenized text data into a language encoder and outputting a text embedding vector, and projecting the text embedding vector into an embedding space through a projection head of the language encoder and projecting the modality embedding vector into the embedding space through a projection head of the vision encoder so that the modality embedding vector and the text embedding vector that correspond to each other are matched.

Devices and techniques are generally described for local device embeddings for automation. In various examples, first data representing first state change data for network-connected computing devices configured in communication with a first network may be determined. The first data may be input into a first machine learning model. In some examples, the first machine learning model may generate first embedding data representing a combination of the first data and second data. In some examples, the second data may represent historical state change data for the network-connected devices. In some examples, the first embedding data may be stored in memory. A first action may be performed by a first network-connected device based at least in part on the first embedding data.

A method and system for cross-modal manifold alignment of different data domains includes determining for a shared embedding space a first embedding function for data of a first domain and a second embedding function for data of a second domain using a triplet loss, wherein triplets of the triplet loss include an anchor data point from the first, a positive and a negative data point from the second domain; creating a first mapping for the data of the first domain using the first embedding function in the shared embedding space; creating a second mapping for the data of the second domain using the second embedding function in the shared embedding space; and generating a cross-modal alignment for the data of the first domain and the data of the second domain.