An image-based method of modeling and rendering a three-dimensional model of an object is provided. The method comprises: obtaining a three-dimensional point cloud at each frame of a synchronized, multi-view video of an object, wherein the video comprises a plurality of frames; extracting a feature descriptor for each point in the point cloud for the plurality of frames without storing the feature descriptor for each frame; producing a two-dimensional feature map for a target camera; and using an anti-aliased convolutional neural network to decode the feature map into an image and a foreground mask.

A method for generating a simplified computer-implemented neural network. The method includes: receiving a predefined neural network, which includes a plurality of neural network structures and is described by weights, each neural network structure being assigned a pruning vector which describes a change in weights as a result of the pruning of the respective neural network; calculating a product of a matrix including a structure vector, the matrix including partial second order derivations of a loss function with respect to the plurality of weights; determining changes in the loss function with respect to the predefined neural network, each change occurring as a result of a pruning of a corresponding neural network structure of the two or more neural network structures to be pruned; and pruning at least one neural network structure based on the determined two or more changes in the loss function to generate the simplified neural network.

Computer-implemented method and system (100) for estimating vegetation coverage in a real-world environment. The system receives an RGB image (91) of a real-world scenery (1) with one or more plant elements (10) of one or more plant species. At least one channel of the RGB image (91) is provided to a semantic regression neural network (120) which is trained to estimate at least a near-infrared channel (NIR) from the RGB image. The system obtains an estimate of the near-infrared channel (NIR) by applying the semantic regression neural network (120) to the at least one RGB channel (91). A multi-channel image (92) comprising at least one of the R-, G-, B-channels (R, G, B) of the RGB image and the estimated near-infrared channel (NIR), is provided as test input (TI1) to a semantic segmentation neural network (130) trained with multi-channel images to segment the test input (TI1) into pixels associated with plant elements and pixels not associated with plant elements. The system segments the test input (TI1) using the semantic segmentation neural network (130) resulting in a vegetation coverage map (93) indicating pixels of the test input associated with plant elements (10) and indicating pixels of the test input not associated with plant elements.

A method for recognizing handwritten text is disclosed. The method comprises receiving data comprising a sequence of ink points; applying the received data to a neural network-based sequence classifier trained with a Connectionist Temporal Classification (CTC) output layer using forced alignment to generate an output; generating a character hypothesis as a portion of the sequence of ink points; applying the character hypothesis to a character classifier to obtain a first probability corresponding to the probability that the character hypothesis includes the given character; processing the output of the CTC output layer to determine a second probability corresponding to the probability that the given character is observed within the character hypothesis; and combining the first probability and the second probability to obtain a combined probability corresponding to the probability that the character hypothesis includes the given character.

The present disclosure describes systems and methods that include calculating, via a reinforcement learning agent (RLA) controller, a plurality of state-action values based on sensor data representing an observed state, wherein the RLA controller utilizes a deep neural network (DNN) and generating, via a fuzzy controller, a plurality of linear models mapping the plurality of state-action values to the sensor data.

Disclosed is a non-transitory computer readable medium storing a computer program. When the computer program is executed by one or more processors of a computing device, the computer program performs the following operations for processing data, and the operations may include: determining an uncertainty level with respect to labeling criteria for each of one or more data included in a dataset; determining a similarity level for one or more data included in a data subset; and selecting at least some of data included in the dataset based on the uncertainty level and the similarity level, and additionally labeling the selected data.

A target location for a therapeutic intervention is determined in a subject with a neurological disorder. The target location is selected within at least one resting state network (RSN) map according to a predetermined criterion for the neurological disorder. The at least one RSN map includes a plurality of functional voxels within a brain of the subject, and each functional voxel of the plurality of functional voxels is associated with a probability of membership in an RSN. Instructions are transmitted to a treatment system that cause operation to be performed on the selected target location.

A system and methods are provided for determining a vehicle action during a phantom projection attack, including processing a received image to identify a traffic object, and creating from the received image multiple processed images that are applied to respective neural network (NN) models. Latent representations of the multiple processed images from each of the NN models are then fed to a combiner model trained to determine whether the latent representations indicate a phantom projection attack, and, responsively to a determination of a phantom projection attack, issuing a phantom projection indicator.

Methods and systems are for improvements to in-vitro fertilization using morpho-kinetic signatures. These improvements are achieved by analyzing a series of images of a developing embryo (e.g., time-lapse images) as opposed to a single static image. For example, due to the difficulty in identifying clear distinctions between morphological states based on static images, as well as the unpredictability of morpho-kinetic development of an embryo, the system analyzes the development of an embryo as a whole over a given time frame (e.g., fertilization to blastulation), which provides a better prediction of the viability of a given embryo. The analysis may take the form of a morpho-kinetic signature, which itself may be used to determine an optimal time to transfer and/or implant an embryo into a patient.

A method for adjusting external air intake in an interior of a vehicle involves continuously identifying an interior pollution level during a driving operation of the vehicle using recorded signals of a pollution sensor arranged in the interior. A pollution level of external air on a section of road ahead of the vehicle is predicted and the external air intake is automatically regulated depending on the predicted pollution level of the external air.