207 results about "Visual search tasks" patented technology

Disclosed is method of visual search for objects that include straight lines. A two-step process is used, which includes detecting straight line segments in an image. The lines are generally characterized by their length, midpoint location, and orientation. Hypotheses that a particular straight line segment belongs to a known object are generated and tested. The set of hypotheses is constrained by spatial relationships in the known objects. The speed and robustness of the method and apparatus disclosed makes it immediately applicable to many computer vision applications.

Each image of a set of images is characterized with a set of sparse feature descriptors and a set of dense feature descriptors. In some embodiments, both the set of sparse feature descriptors and the set of dense feature descriptors are calculated based on a fixed rotation for computing texture descriptors, while color descriptors are rotation invariant. In some embodiments, the descriptors of both sparse and dense features are then quantized into visual words. Each database image is represented by a feature index including the visual words computed from both sparse and dense features. A query image is characterized with the visual words computed from both sparse and dense features of the query image. A rotated local Bag-of-Features (BoF) operation is performed upon a set of rotated query images against the set of database images. Each of the set of images is ranked based on the rotated local Bag-of-Features operation.

A method and apparatus is described here that categorizes images by extracting regions and describing the regions with a 16-dimensional subscene feature vector, which is a concatenation of color, texture, and spatial feature vectors. By comparing the spatial feature vectors in images with similarly-obtained feature vectors in a Gaussian mixture based model pool (obtained in a subscene modeling phase), the images may be categorized (in a subscene recognition phase) with probabilities relating to each region or subscene. Higher probabilities are likelier correlations. The device may be a single or multiple core CPU, or parallelized vector processor for characterizing many images. The images may be photographs, videos, or video stills, without restriction. When used real-time, the method may be used for visual searching or sorting.

Systems, methods, and computer program products for identifying a relevant candidate product in an electronic marketplace. Embodiments perform a visual similarity comparison between candidate product image visual content and input query image visual content, process formal and informal natural language user inputs, and coordinate aggregated past user interactions with the marketplace stored in a knowledge graph. Visually similar items and their corresponding product categories, aspects, and aspect values can determine suggested candidate products without discernible delay during a multi-turn user dialog. The user can then refine the search for the most relevant items available for purchase by providing responses to machine-generated prompts that are based on the initial search results from visual, voice, and / or text inputs. An intelligent online personal assistant can thus guide a user to the most relevant candidate product more efficiently than existing search tools.

Disclosed is method of visual search for objects that include straight lines. A two-step process is used, which includes detecting straight line segments in an image. The lines are generally characterized by their length, midpoint location, and orientation. Hypotheses that a particular straight line segment belongs to a known object are generated and tested. The set of hypotheses is constrained by spatial relationships in the known objects. The speed and robustness of the method and apparatus disclosed makes it immediately applicable to many computer vision applications.

The present invention includes a system and method for generating a discussion in a social network based on visual search results. A mixed media reality (MMR) engine receives images from a user device and identifies matching MMR objects. A social network application determines whether a discussion group that is related to metadata associated with the images from user devices are related to a discussion group. If the discussion group does not yet exist, the social network application generates the discussion group and provides the user devices with information about the discussion group.

Systems, methods, and computer program products for identifying a candidate product in an electronic marketplace based on a visual comparison between candidate product image visual text content and input query image visual text content. Unlike conventional optical character recognition (OCR) based systems, embodiments automatically localize and isolate portions of a candidate product image and an input query image that each contain visual text content, and calculate a visual similarity measure between the respective portions. A trained neural network may be re-trained to more effectively find visual text content by using the localized and isolated visual text content portions as additional ground truths. The visual similarity measure serves as a visual search result score for the candidate product. Any number of images of any number of candidate products may be compared to an input query image to enable text-in-image based product searching without resorting to conventional OCR techniques.

The invention discloses a visual search interface design method provided with the semantic classification function. The design provides a novel information retrieval service; the user can obtain the required key information in the candidate set of the search result conveniently, in particular facilitates the key information browsing of the user. The key implementation steps of the interface design described in the invention description include: the candidate set is obtained through the current commercial search engine; the candidate search result is processed with the semantic classification and the extraction of the subject classification information; in each classification, the selected search result is reported to the user in the summary form and displayed in classifications through the multi-stage or the screen block form. The search interface disclosed by the invention can show all the useful information in the search result set to the user completely, help the user process the network search efficiently and enable the user to quickly obtain the required information from the search result set without browsing the large quantity of repeated information in the search result candidate set.

The present invention is a method and system to permit assessment and measurement of a combination of visual and cognitive processes necessary for safe driving-object detection, by recognition of particular objects as actual or potential safety hazards or traffic hazards, through the continuous switching of visual attention between and among objects while virtually driving, according to an object's instantaneous priority as a safety threats or traffic hazard, that together are termed expert search and scanning skills. The present invention will reference the driver's performance against that of acknowledged experts in this skill; and then provide precise feedback about the type, location, and timing of search and scanning performance errors, plus remedial strategies to correct such errors.

Systems and methods for interactive image compositing may integrate image searching using visual search criteria (e.g., color, composition, tonal distribution, or lighting effects), non-destructive image compositing, and high-performance rendering using a fast gradient solver to provide an efficient workflow for users. An image compositing application, executable by a CPU and / or GPU, may employ a sketch-and-refine approach, whereby a user draws a rough perimeter around an object of interest in an image and (following an initial application of the fast gradient solver to a lower-resolution version of the image) receives a preview of how it would fit into a composition. The fast gradient solver may differentiate between the object and its background by applying artificial intelligence techniques and / or dependent on user-supplied hints. In a second stage, the fast gradient solver may refine the solution for a final rendering of the composite image by operating on successively higher-resolution versions of the image.

A method, apparatus and computer-readable storage medium are provided for providing a list of candidates that may provide information requested by a user. In this regard, an indication of information desired by a user may be received. The indication of the desired information may include media content associated with one or more objects in a real world and / or meta-information providing one or more characteristics associated with the media content, a mobile device and / or the user. A list of candidates for the desired information may then be presented to the user, such as upon a mobile device of the user. In response to an indication of a selection of the user from among the list of candidates or otherwise, an association may be created between the selection of the user and the indication of information desired by the user.

The present invention includes a system and method for generating a discussion in a social network based on visual search results. A mixed media reality (MMR) engine receives images from a user device and identifies matching MMR objects. A social network application determines whether a discussion group that is related to metadata associated with the images from user devices are related to a discussion group. If the discussion group does not yet exist, the social network application generates the discussion group and provides the user devices with information about the discussion group.

The invention relates to a cognition level rehabilitation training system based on an eye tracking technology. The cognition level rehabilitation training system comprises a cognition task presenting module, an infrared eye tracking assembly, an eye movement data analyzing module, a test result feedback module and a stimulation marker guiding module, wherein the cognition task presenting module is used for presenting an attention task, a visual search task, a memory task, a calculation task and a reading task to a testee; the infrared eye tracking assembly is used for acquiring eye movement data of the testee in real time and sending the data to the eye movement data analyzing module; the eye movement data analyzing module is used for evaluating an attention score, a visual search score, a memory score, a calculation score or a reading score of the testee with a human eye vision dynamic analysis method, and sending a result to the test result feedback module and the stimulation marker guiding module; the test result feedback module is used for feeding back the test result to the testee; the stimulation marker guiding module is used for guiding eye movement of the testee.