Georgetown University

Cristopher Flagg

Ophir Frieder

Jian-young Wu

Gholam Motamedi

Technologies are described for reconstructing facial models which are preserved images or images captured from security cameras. The reconstructed models can be three-dimensional (3D) point clouds and can be compared to existing facial models and/or other reconstructed models based on physical geometry. The 3D point cloud models can be encoded into one or more latent space feature vector representations which can allow both local and global geometric properties of a face to be described. The one or more feature vector representations of a target face can be used individually or in combination with other descriptors for recognition, retrieval, and classification tasks. Neural networks can be used in the encoding of the one or more feature vector representations.

Facial recognition using 3D model

Technologies are described for retrieving documents using image representations in the documents and is based on intra-image features. The identification of elements within an image representation can allow for deeper understanding of the image representation and for better relating image representations based on their intra-image features. The intra-image features present in image representations can be used in searches. Search results can further be reranked to improve search results. For example, reranking can allow search results to conform to intra-image dominant image features.

Document retrieval using intra-image relationships

Technologies are described for reconstructing physical objects which are preserved or represented in pictorial records. The reconstructed models can be three-dimensional (3D) point clouds and can be compared to existing physical models and/or other reconstructed models based on physical geometry. The 3D point cloud models can be encoded into one or more latent space feature vector representations which can allow both local and global geometric properties of the object to be described. The one or more feature vector representations of the object can be used individually or in combination with other descriptors for retrieval and classification tasks. Neural networks can be used in the encoding of the one or more feature vector representations.

Document search for document retrieval using 3D model

A gait analysis system, which includes a neural network with a recurrent neural network layer and a fully connected layer, that receives sensor data indicative of an individual's gait and outputs an assessment regarding the individual's health. The neural network is trained using training data indicative of abnormal gaits and normal gaits. To analyze the training data and the sensor data, the recurrent neural network layer parses each piece of data into a series of windows and analyzes each window in series to generate a context vector characterizing each window and the previously analyzed windows. The fully connected layer, having been trained to differentiate between normal gaits and abnormal gaits based on context vectors characterizing the training data, is used to generate a final assessment characterizing the user gate as normal or abnormal using one or more of the context vectors characterizing the sensor data.

Growing community of inventors

Annandale, VA, United States of America

Cristopher Flagg