Abstract: A method for recognizing long-range activities in videos includes segmenting an input video stream to generate multiple frame sets. For each of the frame sets, a frame with a highest likelihood of including one or more actions of a set of predefined actions is identified regardless of its order in the frame set. A global representation of the input stream is generated based on pooled representations of the identified frames. A long-range activity in the video stream is classified based on the global representation.

Abstract: A method for classifying a human-object interaction includes identifying a human-object interaction in the input. Context features of the input are identified. Each identified context feature is compared with the identified human-object interaction. An importance of the identified context feature is determined for the identified human-object interaction. The context feature is fused with the identified human-object interaction when the importance is greater than a threshold.

Abstract: Certain aspects provide a method for determining a solution to a combinatorial optimization problem, including: determining a plurality of subgraphs, wherein each subgraph of the plurality of subgraphs corresponds to a combinatorial variable of the plurality of combinatorial variables; determining a combinatorial graph based on the plurality of subgraphs; determining evaluation data comprising a set of vertices in the combinatorial graph and evaluations on the set of vertices; fitting a Gaussian process to the evaluation data; determining an acquisition function for vertices in the combinatorial graph using a predictive mean and a predictive variance from the fitted Gaussian process; optimizing the acquisition function on the combinatorial graph to determine a next vertex to evaluate; evaluating the next vertex; updating the evaluation data with a tuple of the next vertex and its evaluation; and determining a solution to the problem, wherein the solution comprises a vertex of the combinatorial graph.

Abstract: A dynamic prototype convolution network (DPCN) can achieve sufficient information interaction between support features from a support image and query features from a query image for performing Few-shot semantic segmentation (FSS). A dynamic convolution module (DCM) can generate dynamic filters from a support foreground. Then, information interaction can be achieved by convolution operations over query features, such as by using these dynamic filters. A support activation module (SAM) and a feature filtering module (FFM) can be used to mine context information from a query feature. The SAM can learn to generate a pseudo mask for a query image. The FFM can refine the pseudo mask to filter background information from a query feature. Thus, information both from query and support can be used to achieve more accurate prediction. The DPCN can be used to perform k-shot segmentation.

Abstract: A method of collaboratively training a neural network model, includes receiving a local update from a subset of the multiple users. The local update is related to one or more subsets of a dataset of the neural network model. A local component of the neural network model identifies a subset of the one or more subsets to which a data point belongs. A global update is computed for the neural network model based on the local updates from the subset of the users. The global updates for each portion of the network are aggregated to train the neural network model.

Abstract: A method for object tracking includes receiving a target image of an object of interest. Latent space features of the target image is modified at a forward pass for a neural network by dropping at least one channel of the latent space features, dropping a channel corresponding to a slice of the latent space features, or dropping one or more features of the latent space features. At the forward pass, a location of the object of interest in a search image is predicted based on the modified latent space features. The location of the object of interest is identified by aggregating predicted locations from the forward pass.

Abstract: Aspects described herein provide a method of processing data, including: receiving a set of global parameters for a plurality of machine learning models; processing data stored locally on an processing device with the plurality of machine learning models according to the set of global parameters to generate a machine learning model output; receiving, at the processing device, user feedback regarding machine learning model output for the plurality of machine learning models; performing an optimization of the plurality of machine learning models based on the machine learning output and the user feedback to generate locally updated machine learning model parameters; sending the locally updated machine learning model parameters to a remote processing device; and receiving a set of globally updated machine learning model parameters for the plurality of machine learning models.

Abstract: A method for processing an image is presented. The method locates a subject and an object of a subject-object interaction in the image. The method determines relative weights of the subject, the object, and a context region for classification. The method further classifies the subject-object interaction based on a classification of a weighted representation of the subject, a weighted representation of the object, and a weighted representation of the context region.

Abstract: A method for classifying subject activities in videos includes learning latent (previously generated) concepts that are analogous to nodes of a graph to be generated for an activity in a video. The method also includes receiving video segments of the video. A similarity between the video segments and the previously generated concepts is measured to obtain segment representations as a weighted set of latent concepts. The method further includes determining a relationship between the segment representations and their transitioning pattern over time to determine a reduced set of nodes and/or edges for the graph. The graph of the activity in the video represented by the video segments is generated based on the reduced set of nodes and/or edges. The nodes of the graph are represented by the latent concepts. Subject activities in the video are classified based on the graph.

Abstract: A method for labeling a spherical target includes receiving an input including a representation of an object. The method also includes estimating unconstrained coordinates corresponding to the object. The method further includes estimating coordinates on a sphere by applying a spherical exponential activation function to the unconstrained coordinates. The method also associates the input with a set of values corresponding to a spherical target based on the estimated coordinates on the sphere.

Abstract: Certain aspects provide a method for determining a solution to a combinatorial optimization problem, including: determining a plurality of subgraphs, wherein each subgraph of the plurality of subgraphs corresponds to a combinatorial variable of the plurality of combinatorial variables; determining a combinatorial graph based on the plurality of subgraphs; determining evaluation data comprising a set of vertices in the combinatorial graph and evaluations on the set of vertices; fitting a Gaussian process to the evaluation data; determining an acquisition function for vertices in the combinatorial graph using a predictive mean and a predictive variance from the fitted Gaussian process; optimizing the acquisition function on the combinatorial graph to determine a next vertex to evaluate; evaluating the next vertex; updating the evaluation data with a tuple of the next vertex and its evaluation; and determining a solution to the problem, wherein the solution comprises a vertex of the combinatorial graph.

Abstract: A method of tracking an object includes performing a hybrid search over a sequence of frames. The hybrid search includes periodically performing a global search on selected frames of the sequence of frames and performing a local search on frames between the selected frames of the global search. The method also includes updating a similarity function based on a result of the hybrid search. The method further includes tracking the object based on the hybrid search.

Abstract: An apparatus may be configured to obtain, for a Siamese neural network having a recurrent neural network (RNN), an initial representation associated with a target object at a first time step and a set of candidate regions at a current time step. The apparatus may determine an updated representation associated with the target object based on the initial representation at the first time step and observed information associated with the target object at a set of previous time steps, and the observed information associated with the target object may be represented by a hidden state of the RNN. The apparatus may output the updated representation associated with the target object for matching with the set of candidate regions at the current time step by the Siamese neural network. The apparatus may determine the updated representation further based on a hidden state at a previous time step.

Abstract: A method for classifying subject activities in videos includes learning latent (previously generated) concepts that are analogous to nodes of a graph to be generated for an activity in a video. The method also includes receiving video segments of the video. A similarity between the video segments and the previously generated concepts is measured to obtain segment representations as a weighted set of latent concepts. The method further includes determining a relationship between the segment representations and their transitioning pattern over time to determine a reduced set of nodes and/or edges for the graph. The graph of the activity in the video represented by the video segments is generated based on the reduced set of nodes and/or edges. The nodes of the graph are represented by the latent concepts. Subject activities in the video are classified based on the graph.

Abstract: A method for processing an image is presented. The method locates a subject and an object of a subject-object interaction in the image. The method determines relative weights of the subject, the object, and a context region for classification. The method further classifies the subject-object interaction based on a classification of a weighted representation of the subject, a weighted representation of the object, and a weighted representation of the context region.

Abstract: A method aligns, with an artificial neural network, a three-dimensional (3D) model to an object in a 2D image. The method includes detecting, with an object detector, the object from the 2D image. The method also includes estimating a geodesic distance value between the object and multiple discretized poses of the 3D model. The method further includes selecting a discretized pose of the multiple discretized poses corresponding to a smallest geodesic distance value. The method still further includes propagating pose parameters of the selected discretized pose of the 3D model to the object.

Abstract: A method of tracking a position of a target object in a video sequence includes identifying the target object in a reference frame. A generic mapping is applied to the target object being tracked. The generic mapping is generated by learning possible appearance variations of a generic object. The method also includes tracking the position of the target object in subsequent frames of the video sequence by determining whether an output of the generic mapping of the target object matches an output of the generic mapping of a candidate object.

Abstract: A method for labeling a spherical target includes receiving an input including a representation of an object. The method also includes estimating unconstrained coordinates corresponding to the object. The method further includes estimating coordinates on a sphere by applying a spherical exponential activation function to the unconstrained coordinates. The method also associates the input with a set of values corresponding to a spherical target based on the estimated coordinates on the sphere.

Abstract: A method of tracking a position of a target object in a video sequence includes identifying the target object in a reference frame. A generic mapping is applied to the target object being tracked. The generic mapping is generated by learning possible appearance variations of a generic object. The method also includes tracking the position of the target object in subsequent frames of the video sequence by determining whether an output of the generic mapping of the target object matches an output of the generic mapping of a candidate object.

Abstract: A method, a computer-readable medium, and an apparatus for zero-exemplar event detection are provided. The apparatus may receive a plurality of text blocks, each of which may describe one of a plurality of pre-defined events. The apparatus may receive a plurality of training videos, each of which may be associated with one of the plurality of text blocks. The apparatus may propagate each text block through a neural network to obtain a textual representation in a joint space of textual and video representations. The apparatus may propagate each training video through the neural network to obtain a visual representation in the joint space. The apparatus may adjust parameters of the neural network to reduce, for each pair of associated text block and training video, the distance in the joint space between the textual representation of the associated text block and the visual representation of the associated training video.