Huawei Technologies Co., Limited

Zhixiang Chi

Jin Tang

Yuanhao Yu

Yang Wang

Zheng Liu

Rasoul Mohammadi Nasiri

Juwei Lu

Huan Liu

Li Gu

Systems, methods and computer-readable medium for predicting a depth for a video frame are disclosed. An example method may include steps of: receiving a plurality of training data, each comprising a set of consecutive video frames and a depth representation of a subsequent video frame to the consecutive video frames; receiving a pre-trained neural network model fhaving a plurality of weights θ; while the pre-trained neural network model fhas not converged: computing a plurality of second weights θ', based on each set of consecutive video frames, and updating the plurality of weights θ, based on the plurality of training data and the plurality of second weights θ′; receiving a plurality of new consecutive video frames with consecutive timestamps; and predicting a depth representation of video frame immediately subsequent to the new consecutive video frames based on the updated plurality of weights θ.

Methods, systems and computer medium for scene-adaptive future depth prediction in monocular videos

This disclosure provides for methods and system for meta few-shot class incremental learning. According to an aspect a method is provided. The method includes obtaining at least one weight attention map of a first network and updating weights of a second network using the at least one weight attention map, where the second network is a modulatory network. The method further includes generating at least one feature attention map of the second network based on the at least one weight attention map of the first network and a set of input images of at least one class. The method further includes generating at least one feature map of the first network based on the set of input images of the at least one class, and updating the at least one feature map of the first network based on the feature attention map of the second network.

Meta few-shot class incremental learning

Methods and systems for image deblurring are described. The weights of a deblurring network are first trained by meta-training the deblurring network on both a primary deblurring task and an auxiliary reconstruction task. Application-time training of the deblurring network is then performed, using an application-time blurry input image, to obtain values of application-time trained weights. Application-time training includes performing the auxiliary reconstruction task on the application-time blurry input image, and updating the weights of the deblurring network based on an auxiliary loss computed from the auxiliary reconstruction task. A deblurred output image is generated from the application-time blurry input image, using the application-time trained weights in the deblurring network.

Methods and systems for deblurring blurry images

Systems and methods for multi-frame video frame interpolation. Higher-order motion modeling, such as cubic motion modeling, achieves predictions of intermediate optical flow between multiple interpolated frames, assisted by relaxation of the constraints imposed by the loss function used in initial optical flow estimation. A temporal pyramidal optical flow refinement module performs coarse-to-fine refinement of the optical flow maps used to generate the intermediate frames, focusing a proportionally greater amount of refinement attention to the optical flow maps for the high-error middle frames. A temporal pyramidal pixel refinement module performs coarse-to-fine refinement of the generated intermediate frames, focusing a proportionally greater amount of refinement attention to the high-error middle frames. A generative adversarial network (GAN) module calculates a loss function for training the neural networks used in the optical flow estimation module, temporal pyramidal optical flow refinement module, and/or temporal pyramidal pixel refinement module.

Systems and methods for multi-frame video frame interpolation

A system, device and method of video frame interpolation. In one aspect, a frame interpolation system receive candidate interpolated frame and difference data based on a first frame and second frame. It identifies a high-error region and one or more increasingly lower-error regions of the candidate interpolated frame based on the difference data and applies multiple successive stages of a refinement neural network to the regions of the candidate interpolated frame, from highest- to lowest-error. At each stage, the results of the previous stage are combined with the next lower-error region before being refined by the current stage. In some aspects, the system also generates the candidate interpolated frame and the difference data using a flow estimation neural network.

Growing community of inventors

North York, Canada

Zhixiang Chi