Abstract: Systems, methods, and computer media of processing a video are disclosed. An example method may include: receiving a plurality of video frames of a video; generating a plurality of first input features based on the plurality of video frames; generating a plurality of second input features based on reversing a temporal order of the plurality of first input features; generating a first set of joint attention features based on the plurality of first input features; generating a second set of joint attention features based on the plurality of second input features; and concatenating the first set of joint attention features and the second set of joint attention features to generate a final set of joint attention features.

Abstract: Methods and systems are described for performing video retrieval together with video grounding. A word-based query for a video is and encoded into a query representation using a trained query encoder. One or more similar video representations are identified, from a plurality of video representations that are similar to the query representation. Each similar video representation represents a respective relevant video. A grounding is generated for each relevant video by forward propagating each respective similar video representation together with the query representation through a trained grounding module. The relevant videos or identifiers of the relevant videos are outputted together with the grounding generated for each relevant video.

Abstract: An adaptive action recognizer for video that performs multiscale spatiotemporal decomposition of video to generate lower complexity video. The adaptive action recognizer has a number of processing pathways, one for each level of video complexity with each processing pathway having a different computational cost. The adaptive action recognizer applies a decision making scheme that encourages using low average computational costs while retaining high accuracy.

Abstract: Methods and systems are described for performing video retrieval together with video grounding. A word-based query for a video is and encoded into a query representation using a trained query encoder. One or more similar video representations are identified, from a plurality of video representations that are similar to the query representation. Each similar video representation represents a respective relevant video. A grounding is generated for each relevant video by forward propagating each respective similar video representation together with the query representation through a trained grounding module. The relevant videos or identifiers of the relevant videos are outputted together with the grounding generated for each relevant video.

Abstract: Provided are methods for unsupervised domain adaptation for LiDAR segmentation via enhanced pseudo-labelling techniques, which can include training a machine learning model to perform a segmentation task for a source domain using a first sample set. Some methods also include generating a second sample set by applying the trained model to one or more unannotated samples associated with a target domain, and annotating the one or more unannotated samples with one or more pseudo-labels corresponding to an output of the trained machine learning model. Some methods also include generating a third sample set that includes at least one sample formed by concatenating a first sample from the first sample set and a second sample from the second sample set with target inputs. Some methods also include updating the trained machine learning model to perform the segmentation task for the target domain. Systems and computer program products are also provided.

Abstract: Systems, methods, and computer media of processing a video are disclosed. An example method may include: receiving a plurality of video frames of a video; generating a plurality of first input features based on the plurality of video frames; generating a plurality of second input features based on reversing a temporal order of the plurality of first input features; generating a first set of joint attention features based on the plurality of first input features; generating a second set of joint attention features based on the plurality of second input features; and concatenating the first set of joint attention features and the second set of joint attention features to generate a final set of joint attention features.

Abstract: Methods and systems for updating the weights of a set of convolution kernels of a convolutional layer of a neural network are described. A set of convolution kernels having attention-infused weights is generated by using an attention mechanism based on characteristics of the weights. For example, a set of location-based attention multipliers is applied to weights in the set of convolution kernels, a magnitude-based attention function is applied to the weights in the set of convolution kernels, or both. An output activation map is generated using the set of convolution kernels with attention-infused weights. A loss for the neural network is computed, and the gradient is back propagated to update the attention-infused weights of the convolution kernels.

Abstract: Methods, processing units and media for multi-bandwidth separated feature extraction convolution in a neural network are described. A convolution block splits input channels of an activation map into multiple branches, each branch undergoing convolution at a different bandwidth by using down-sampling of the inputs. The outputs are concatenated by up-sampling the outputs of the low-bandwidth branches using pixel shuffling. The concatenation operation may be a shuffled concatenation operation that preserves separated multi-bandwidth feature information for use by subsequent layers of the neural network. Embodiments are described which apply frequency-based and magnitude-based attention to the weights of the convolution kernels based on the frequency band locations of the weights.

Abstract: Methods and systems for updating the weights of a set of convolution kernels of a convolutional layer of a neural network are described. A set of convolution kernels having attention-infused weights is generated by using an attention mechanism based on characteristics of the weights. For example, a set of location-based attention multipliers is applied to weights in the set of convolution kernels, a magnitude-based attention function is applied to the weights in the set of convolution kernels, or both. An output activation map is generated using the set of convolution kernels with attention-infused weights. A loss for the neural network is computed, and the gradient is back propagated to update the attention-infused weights of the convolution kernels.

Abstract: An adaptive action recognizer for video that performs multiscale spatiotemporal decomposition of video to generate lower complexity video. The adaptive action recognizer has a number of processing pathways, one for each level of video complexity with each processing pathway having a different computational cost. The adaptive action recognizer applies a decision making scheme that encourages using low average computational costs while retaining high accuracy.