1668results about "Still image data querying" patented technology

Systems and techniques are provided for detecting gaze direction of subjects in an area of real space. The system receives a plurality of sequences of frames of corresponding fields of view in the real space. The system uses sequences of frames in a plurality of sequences of frames to identify locations of an identified subject and gaze directions of the subject in the area of real space over time. The system includes logic having access to a database identifying locations of items in the area of real space. The system identifies items in the area of real space matching the identified gaze directions of the identified subject.

In an embodiment, a method and apparatus for generating a presentation is provided. The method considers characteristics of audio works and visual works when constructing the presentation. In some embodiments, the presentation may be automatically constructed.

The invention discloses an image retrieval method based on multi-task hash learning. Firstly, the deep convolutional neural network model is determined. Secondly, the loss function is designed by using multi-task learning mechanism. Then, the training method of convolutional neural network model is determined, in combination with the loss function, and back propagation method is used to optimize the model. Finally, the image is input to the convolutionalal neural network model, and the output of the model is transformed into hash code for image retrieval. The convolutional neural network modelis composed of a convolutional sub-network and a full connection layer. The convolutional subnetwork consists of a first convolutional layer, a maximum pooling layer, a second convolutional layer, anaverage pooling layer, a third volume base layer and a spatial pyramid pooling layer. The full connection layer is composed of a hidden layer, a hash layer and a classification layer. The training method of the model includes two training methods: a combined training method and a separated training method. The method of the invention can effectively retrieve single tag and multi-tag images, and the retrieval performance is better than other deep hashing methods.

The invention is related to a method of matching image features with reference features, comprising the steps of providing a current image captured by a capturing device, providing reference features (r), wherein each of the reference features comprises at least one reference feature descriptor (d(r)), determining current features (c) in the current image and associating with each of the current features at least one respective current feature descriptor (d(c)), and matching the current features with at least some of the reference features by determining a respective similarity measure (D(c, r)) between each respective current feature descriptor (d(c)) and each respective reference feature descriptor (d(r)). According to the invention, the determination of the similarity measure is performed on an integrated circuit by hardwired logic or configurable logic which processes logical functions for determining the similarity measure. The invention is also concerned with an integrated circuit for matching of image features with reference features.

System and method for creating a collection of images are described, the method comprising: receiving images from at least one source of images; processing the images to produce an output collection of images, the processing comprising grouping the images to clusters of related images and selecting the preferred images in the clusters; and outputting the output collection of images, the output collection of images comprising the clusters of related images and indication of the preferred images in the clusters. The system for creating a collection of images comprising: a storage medium to receive images from at least one source of images; a processor to produce an output collection of images by grouping the images to clusters of related images and selecting the preferred images in the clusters; and a collection output medium for outputting the output collection of images.

The invention provides a method and a system for assisting a teacher to understand the learning situation of a student. The method comprises the following steps: timing is started when a student starts to answer a question in each answer area, the timing is stopped until the answer is stopped in each answer area; Shooting and acquiring an answer image corresponding to each answer area after stopping answering in each answer area; Identifying the answer image, comparing the recognized answer content with the preset answer to obtain the corresponding correction result; According to each answer area corresponding to the correction results and answer time, obtaining the students' learning situation. The invention not only detaches from electronic equipment such as mobile phone and tablet computer, but also avoids students playing mobile phone or tablet computer on the grounds of learning, thus delaying learning. The invention is more in line with the student 's learning environment, assists in correcting students' homework and improves the efficiency of correcting homework. The accurate grasp of each student 's learning situation is conducive to the follow-up of each student' s accurate differentiation training, and to improving students' performance.

Apparatuses and methods are provided for storing information related to objects associated with a hand of a user via a wearable camera system. In one implementation, a wearable apparatus for storing the information is provided comprising a wearable image sensor configured to capture a plurality of images from the environment of the user, and at least one processing device programmed to process the images. The processing device may detect the hand of the user, and an object associated with the user's hand. The processing device may proceed to store information related to the object. Consistent with disclosed embodiments, the stored information may be used for various purposes, such as warning the user of dangers, catering advertising to the user, and helping the user find objects when they are lost.

A method of identifying one or more particular images from an image collection, includes indexing the image collection to provide image descriptors for each image in the image collection such that each image is described by one or more of the image descriptors; receiving a query from a user specifying at least one keyword for an image search; and using the keyword(s) to search a second collection of tagged images to identify co-occurrence keywords. The method further includes using the identified co-occurrence keywords to provide an expanded list of keywords; using the expanded list of keywords to search the image descriptors to identify a set of candidate images satisfying the keywords; grouping the set of candidate images according to at least one of the image descriptors, and selecting one or more representative images from each grouping; and displaying the representative images to the user.

The invention discloses an audio processing method and device based on artificial intelligence. One concrete implement mode of the method includes the steps that an audio file to be processed is converted into an image to be processed; the content feature of the image to be processed is extracted; according to the style feature and the content feature of the image to be processed, a target image is determined, and the style feature is obtained from a template image converted from a template audio file; the target image is converted into the processed audio file. By means of the implement mode, the processed audio file has the template audio style without changing the content of the audio file to be processed, and audio processing efficiency and flexibility are improved.

A method of providing an electronic filing system which is searchable using a handwritten search query, the method including the steps of: obtaining the handwritten search query using an input device; performing a search of at least one database based on a comparison between the handwritten search query and handwritten annotations made on interactive pages stored in the at least one database; and, providing the results of the search to a user and facilitating access to at least one interactive page identified in the results of the search. An apparatus is also disclosed. Preferably, the interactive page is provided to the user in the form of printed paper, and the handwritten annotations are user handwriting, symbols, drawings, indicia or the like.

The subject matter described herein includes methods, systems, and computer program products for resource-to-resource metadata association. According to one method, a source resource and a destination resource are identified. A type of at least one of the source and destination resources is identified and used to select a transform for mapping data values associated with the resource to the destination resource as metadata. Based on the transform, at least one of the data values associated with the source resource is associated with a destination resource as metadata.

The invention belongs to the technical field of cross-modal retrieval, and particularly relates to an image-text retrieval system and method based on a multi-angle self-attention mechanism. The systemcomprises a deep convolutional network, a bidirectional recurrent neural network, an image, a text self-attention network, a multi-modal space mapping network and a multi-stage training module. The deep convolutional network is used for acquiring an embedding vector of an image region feature in an image embedding space. The bidirectional recurrent neural network is used for acquiring an embedding vector of a word feature in a text space, and the two vectors are respectively input to the image and the text self-attention network. The image and text self-attention network is used for acquiringan embedded representation of an image key area and an embedded representation of key words in sentences. The multi-modal space mapping network is used for acquiring the embedded representation of the image text in the multi-modal space. The multi-stage training module is used for learning parameters in the network. A good result is obtained on a common data set Flickr30k and an MSCOCO, and the performance is greatly improved.

The invention provides a zero sample sketch retrieval method based on a semantic adversarial network, which mainly solves the problems that in the prior art, the sketch intra-class variance is larger,and the visual knowledge is difficult to migrate from a known class to a non-seen class under the zero sample setting. The method comprises the steps of obtaining a training sample set, constructinga semantic adversarial network, and extracting the RGB image features through a VGG16 network, constructing a generation network to generate the RGB image features with discriminability, inputting theto-be-retrieved sketch into a semantic confrontation network to generate the semantic features, inputting the semantic features and the random Gaussian noise into the generation network to generate the RGB image features, and searching the first 200 images most similar to the RGB image features in an image retrieval library to obtain a retrieval result. According to the method, the intra-class variance of the sketch image features is reduced, the RGB image features generated according to the sketch image in each class can be ensured, the retrieval performance of zero sample sketch retrieval is improved, and the method can be used for the electronic commerce, medical diagnosis and remote sensing imaging.

The invention discloses an image data processing method, an apparatus, a computer device and a storage medium, which are applied in the technical field of image recognition. The method comprises the following steps: crawling an original image by a crawler tool, wherein each original image corresponds to an image type; at least one text line region is obtain by using a text position algorithm to locate that original image, and each text line region is screened to obtain a region block image; based on the image type and the position information of the region block image, the target OCR recognition model is obtained. A target OCR recognition model is adopted to recognize a region block image, and a target recognition result is obtained, the target recognition result comprises at least two recognized characters and a corresponding recognition probability; a target text is obtained based on at least two recognize characters and corresponding recognition probability, that target text is determined as a labeled text, and a target image sample is obtain based on a region block image and a labeled text. The method can effectively improve the acquisition efficiency of target image samples and reduce the acquisition cost.

The invention provides a file creation method, a device, a device and a storage medium, belonging to the technical field of image processing. The file building method comprises the following steps: clustering the captured portrait to obtain at least one clustering result; Determining a standard portrait matching the clustering result from the archival database; filing the clustering result with the standard portrait. The invention aggregates the portraits of the same person at different times or at different positions in a clustering result by performing clustering treatment on the portraits captured, a standard portrait matching the clustering result is determined from the file database; The clustering result is filed in the personal file corresponding to the standard portrait, The invention realizes the file creation of the captured portrait, saves manpower to a great extent, effectively improves the accuracy of the archives establishment, solves the work of analyzing and archiving the number of the portrait at the tens of billions level, ensures the accuracy of the archives establishment, and can be competent for large-scale file management.

The invention discloses an image retrieval method, and the method comprises the following steps: obtaining a to-be-retrieved target image, and inputting the target image into a target deep learning model; utilizing the target deep learning model to perform feature extraction on the target image to obtain image features of the target image; wherein the image features comprise global features, localfeatures and multi-scale global features, and the multi-scale global features are features obtained by performing weighted calculation on a plurality of intermediate-stage features generated in the global feature extraction process; respectively calculating similar distances between the target image and the images in the image library by utilizing the image features according to a distance calculation rule; and determining and outputting a similar image of the target image by using the similar distance. According to the method, the image retrieval accuracy can be improved. The invention further discloses an image retrieval device and equipment and a readable storage medium which have corresponding technical effects.

The invention relates to an expression image recommendation method and device based on voice emotion recognition, an electronic device and a storage medium. The method comprises the following steps ofacquiring a plurality of latest voice messages in a current interaction window of the instant messaging software, and extracting audio feature vectors of the voice messages; matching the audio feature vector of the voice information with a plurality of emotion feature models, wherein the plurality of emotion feature models respectively correspond to one of a plurality of emotion classifications;taking the emotion classification corresponding to the matched emotion feature model as the emotion classification of the voice information; and determining one or more target expression images basedon the emotion classification of the voice information, and recommending the target expression images to the current user, so that the efficiency of selecting the expression image is improved.

A visual query system, and associated method and computer program product enhance and accelerate content-based querying, and present a new image similarity measure using known or available software applications and hardware components of video compression systems. The present system encodes images as consecutive frames in a video sequence and uses the ratio between the file length of the compressed sequence and the original file length as a distance measure. The system considers the compression ratio to be an estimate of the entropy of the combined images, which can be used to estimate the amount of new information introduced from one image to the other.

The embodiment of the invention discloses an image-text matching method and device, a storage medium and equipment, and belongs to the technical field of computers. The image-text matching method comprises the steps of obtaining an image and a text to be matched; generating a candidate instance feature set according to the image; aggregating candidate instance features in the candidate instance feature set by using a self-attention mechanism to obtain an instance feature set, each instance feature in the instance feature set corresponding to one object in the image; encoding the text to obtaina text vector; and calculating a matching degree between the image and the text according to the instance feature set and the text vector. According to the embodiment of the invention, the realization difficulty of image-text matching can be simplified, and the accuracy of image-text matching can be improved.

The invention discloses a small sample learning algorithm based on covariance measurement, and belongs to the field of computer vision. The algorithm comprises the steps: (1) introducing an interpolation training mechanism to learn migratable knowledge; (2) designing a local covariance representation, and then embedding the local covariance representation into a deep network to learn and express each concept; (3) constructing a covariance measurement layer to measure the distribution consistency between a query sample and the concepts based on the local covariance representation. According tothe method, a novel and concise end-to-end covariance measurement network CovaMNet is provided, a second-order local covariance representation is designed to replace traditional first-order concept representation, and a new covariance measurement function is provided. A comparison experiment result on a plurality of reference data sets is analyzed to obtain that the CovaMNet framework provided bythe invention shows a competitive effect on a general small sample classification task and a fine-grained small sample classification task.

The invention relates to a video image data retrieval method and device, an apparatus and a storage medium. The method comprises the steps of obtaining a picture retrieval database and a training database; performing clustering training on the feature data in the training database to generate a preset number of data buckets, and determining a clustering center of each data bucket; calculating thedistance between each piece of feature data in the picture retrieval database and each clustering center, and adding each piece of feature data in the picture retrieval database into the correspondingdata bucket according to a first distance rule to determine an inverted index table; calculating the distance between the feature matrix of the to-be-retrieved picture and each clustering center, anddetermining a target data bucket according to a second distance rule; and based on the inverted index table, calculating the distance between the feature vector matrix of the to-be-retrieved pictureand the clustering center of the target data bucket, and determining a picture similar to the to-be-retrieved picture as a retrieval result according to a retrieval rule, so that the performance and the efficiency of the video image data retrieval are improved.

A wearable apparatus and method are provided for selectively categorizing information derived from images. In one implementation, a wearable apparatus for collecting information related to activities of a user may collect information related to activities of a user. The wearable apparatus may include an image sensor configured to capture a plurality of images from an environment of a user; a communications interface; and at least one processing device. The at least one processing device may be programmed to process the plurality of images to identify an activity occurring in the environment of the user; associate the activity with an activity category; and cause transmission of at least the activity category to a remotely located computing device via the communications interface.