Abstract: An indexed hierarchical tree search structure converts each registration sample into an equivalent registration model based on the clustering of its registration item descriptors in the leaf nodes of the hierarchical tree. Query item descriptors from a query sample from someone wanting to be recognized are distributed into the hierarchical tree. A query model is defined based on the clustering of query item descriptors at the leaf nodes, and registration and verification are made based on comparison of the query model and the registration models.

Abstract: Embodiments of the present invention include systems and methods for improved state space modeling (SSM) comprising two added layers to model the substructure transition dynamics and action duration distribution. In embodiments, the first layer represents a substructure transition model that encodes the sparse and global temporal transition probability. In embodiments, the second layer models the action boundary characteristics by injecting discriminative information into a logistic duration model such that transition boundaries between successive actions can be located more accurately; thus, the second layer exploits discriminative information to discover action boundaries adaptively.

Abstract: An echo canceller and an echo cancellation method are provided. The echo canceller includes a self-adaptive filter, a voice signal detection portion and a path change detection portion; a far-end voice signal is propagated in an echo path through a speaker and is picked up by a microphone to form an echo signal. The self-adaptive filter is configured to receive the far-end voice signal as a training signal to simulate the echo path, and cancel the echo signal in a near-end signal. The voice signal detection portion is configured to detect a communication status, control the self-adaptive filter according to the communication status, and control startup of the path change detection portion according to the communication status. The path change detection portion is configured to detect whether a change occurs on the echo path, and control the self-adaptive filter according to whether the change occurs on the echo path.

Abstract: Aspects of the present invention include point set matching systems and methods. In embodiments, a tree model is used to find candidate matching locations for a set of query points. In embodiments, a similitude transform is assumed, and the parameters are separately solved to reduce computation complexity. In embodiments, the dominant scaling (?) and rotation (R) parameters are obtained by identifying a maximum in an accumulator space. A translation (t) matrix is calculated in another 1D accumulator space. With the obtained similitude transform, outliers can be reliably detected. This two-stage approach reduces the complexity and calculation time of determining a similitude transform and increases the accuracy and ability to detect outliers.

Abstract: Embodiments of the present invention include systems and methods for identifying an object in an image. In embodiments, object identification includes using smooth encoding from a tree structure to generate a feature from a descriptor. In embodiments, the smooth encoding may be performed by, having identifying a leaf node for a descriptor, moving up the tree voting structure a number of levels from the identified leaf node to a branch node to identify leaf nodes dependent from the branch node; and then, by determining a sparse code under a condition that a distance between the descriptor and centroids of the leaf nodes dependent from the branch node weighted by the sparse code is minimized, wherein each element of the sparse code representing a weight corresponding to leaf nodes.

Abstract: Embodiments of the present invention include systems and methods for identifying and locating an object in an image. In embodiments, an object in an image may be identified by segmenting a first image of an object into one or more superpixels; extracting local descriptors from the first image, each of the descriptors having an interest point with a location; correlating the local descriptors to the superpixels based on locations of the local descriptors and superpixels; determining a probability for an object label for each of a set of the superpixels; and assigning an object label to each of the set of the superpixels based on the probability and a smoothness factor that includes weighting in terms of one or more of spatial, colors, angular distances between superpixels. The superpixels of an image may be concatenated to predict an object label for the image and to determine the location of the image.

Abstract: Aspects of the present invention include systems and methods for generating an optical flow-based feature. In embodiments, to extract an optical flow feature, the optical flow at sparse interest points is obtained, and Locality-constrained Linear Coding (LLC) is applied to the sparse interest points to embed each flow into a higher-dimensional code. In embodiments, for an image frame, the multiple codes are combined together using a weighted pooling that is related to the distribution of the optical flows in the image frame. In embodiments, the feature may be used in training models to detect actions, in trained models for action detection, or both.

Abstract: An automated registration and authentication system combines a generative and discriminative approach to improve the matching of a query object to a database of registered objects. The discriminative approach uses a voting mechanism to identify a most likely match, and the generative approach uses ASIFT transforms to determine a best geometric match. The two results are combined using a technique base on Bayesian inference theory.

Abstract: Embodiments of the present invention include systems and methods for identifying an object in an image. In embodiments, object identification includes using smooth encoding from a tree structure to generate a feature from a descriptor. In embodiments, the smooth encoding may be performed by, having identifying a leaf node for a descriptor, moving up the tree voting structure a number of levels from the identified leaf node to a branch node to identify leaf nodes dependent from the branch node; and then, by determining a sparse code under a condition that a distance between the descriptor and centroids of the leaf nodes dependent from the branch node weighted by the sparse code is minimized, wherein each element of the sparse code representing a weight corresponding to leaf nodes.

Abstract: Aspects of the present invention include point set matching systems and methods. In embodiments, a tree model is used to find candidate matching locations for a set of query points. In embodiments, a similitude transform is assumed, and the parameters are separately solved to reduce computation complexity. In embodiments, the dominant scaling (?) and rotation (R) parameters are obtained by identifying a maximum in an accumulator space. A translation (t) matrix is calculated in another 1D accumulator space. With the obtained similitude transform, outliers can be reliably detected. This two-stage approach reduces the complexity and calculation time of determining a similitude transform and increases the accuracy and ability to detect outliers.

Abstract: Embodiments of the present invention include systems and methods for identifying and locating an object in an image. In embodiments, an object in an image may be identified by segmenting a first image of an object into one or more superpixels; extracting local descriptors from the first image, each of the descriptors having an interest point with a location; correlating the local descriptors to the superpixels based on locations of the local descriptors and superpixels; determining a probability for an object label for each of a set of the superpixels; and assigning an object label to each of the set of the superpixels based on the probability and a smoothness factor that includes weighting in terms of one or more of spatial, colors, angular distances between superpixels. The superpixels of an image may be concatenated to predict an object label for the image and to determine the location of the image.

Abstract: An indexed hierarchical tree search structure converts each registration sample into an equivalent registration model based on the clustering of its registration item descriptors in the leaf nodes of the hierarchical tree. Query item descriptors from a query sample from someone wanting to be recognized are distributed into the hierarchical tree. A query model is defined based on the clustering of query item descriptors at the leaf nodes, and registration and verification are made based on comparison of the query model and the registration models.

Abstract: An indexed hierarchical tree search structure implements incremental tree modification to add new registrants to the tree without requiring reconstruction of the tree. Only data within a leaf node that receives new data is used in the incremental tree modification. Consequently, the original data set used in the creation of the hierarchical tree may be deleted after construction of the tree.

Abstract: An automated registration and authentication system combines a generative and discriminative approach to improve the matching of a query object to a database of registered objects. The discriminative approach uses a voting mechanism to identify a most likely match, and the generative approach uses ASIFT transforms to determine a best geometric match. The two results are combined using a technique base on Bayesian inference theory.

Abstract: Systems and methods are disclosed for generating super resolution images by building a set of multi-resolution bases from one or more training images; estimating a sparse resolution-invariant representation of an image, and reconstructing one or more missing patches at any resolution level.

Abstract: Embodiments of the present invention include systems and methods for improved state space modeling (SSM) comprising two added layers to model the substructure transition dynamics and action duration distribution. In embodiments, the first layer represents a substructure transition model that encodes the sparse and global temporal transition probability. In embodiments, the second layer models the action boundary characteristics by injecting discriminative information into a logistic duration model such that transition boundaries between successive actions can be located more accurately; thus, the second layer exploits discriminative information to discover action boundaries adaptively.

Abstract: Systems and methods are disclosed to perform image parsing on one or more images by identifying a set of similar regions from each image; assigning one or more region labels to each region and generating multiple hypotheses for region label assignment; and detecting class, location and boundary of each object in the image, wherein object classification, detection and segmentation are performed jointly during image parsing.

Abstract: Embodiments of the disclosed technology provide a detecting device for recovering of release film and a recovery apparatus for release film. The recovery apparatus comprises clips for gripping release film, the detecting device comprises: a detecting sensor; a mobile module coupled onto one clipping arm of the clip. When the release film is gripped by the clip, the release film shifts the mobile module to a first position where the mobile module portion enters the detection scope of the detecting sensor; when the release film is not gripped by the clip, the mobile module shifts to a second position, where the whole mobile module portion withdraws out of the detection scope of the detecting sensor.

Abstract: An earpiece for communications comprises a frame, a vibration system disposed in the frame and a magnetic field system disposed in the frame. The vibration system includes a vibrating armature, a diaphragm and a connecting rod. The magnetic field system includes a magnet to generate a first magnetic field on a first axis and a coil to generate a second magnetic field on a second axis. One of the magnet and the coil is connected to the frame. The first and the second magnetic fields produce a net magnetic force to cause the vibration system to move. The first axis is parallel to the second axis.

Abstract: Aspects of the present invention include systems and methods for segmentation and recognition of action primitives. In embodiments, a framework, referred to as the Continuous Linear Dynamic System (CLDS), comprises two sets of Linear Dynamic System (LDS) models, one to model the dynamics of individual primitive actions and the other to model the transitions between actions. In embodiments, the inference process estimates the best decomposition of the whole sequence into continuous alternating between the two set of models, using an approximate Viterbi algorithm. In this way, both action type and action boundary may be accurately recognized.