Abstract: A method, system and computer program product for segmenting generic foreground objects in images and videos. For segmenting generic foreground objects in videos, an appearance stream of an image in a video frame is processed using a first deep neural network. Furthermore, a motion stream of an optical flow image in the video frame is processed using a second deep neural network. The appearance and motion streams are then joined to combine complementary appearance and motion information to perform segmentation of generic objects in the video frame. Generic foreground objects are segmented in images by training a convolutional deep neural network to estimate a likelihood that a pixel in an image belongs to a foreground object. After receiving the image, the likelihood that the pixel in the image is part of the foreground object as opposed to background is then determined using the trained convolutional deep neural network.

Abstract: A method, system and computer program product for segmenting generic foreground objects in images and videos. For segmenting generic foreground objects in videos, an appearance stream of an image in a video frame is processed using a first deep neural network. Furthermore, a motion stream of an optical flow image in the video frame is processed using a second deep neural network. The appearance and motion streams are then joined to combine complementary appearance and motion information to perform segmentation of generic objects in the video frame. Generic foreground objects are segmented in images by training a convolutional deep neural network to estimate a likelihood that a pixel in an image belongs to a foreground object. After receiving the image, the likelihood that the pixel in the image is part of the foreground object as opposed to background is then determined using the trained convolutional deep neural network.

Abstract: A method, system and computer program product for segmenting generic foreground objects in images and videos. For segmenting generic foreground objects in videos, an appearance stream of an image in a video frame is processed using a first deep neural network. Furthermore, a motion stream of an optical flow image in the video frame is processed using a second deep neural network. The appearance and motion streams are then joined to combine complementary appearance and motion information to perform segmentation of generic objects in the video frame. Generic foreground objects are segmented in images by training a convolutional deep neural network to estimate a likelihood that a pixel in an image belongs to a foreground object. After receiving the image, the likelihood that the pixel in the image is part of the foreground object as opposed to background is then determined using the trained convolutional deep neural network.

Abstract: A method, system and computer program product for segmenting generic foreground objects in images and videos. For segmenting generic foreground objects in videos, an appearance stream of an image in a video frame is processed using a first deep neural network. Furthermore, a motion stream of an optical flow image in the video frame is processed using a second deep neural network. The appearance and motion streams are then joined to combine complementary appearance and motion information to perform segmentation of generic objects in the video frame. Generic foreground objects are segmented in images by training a convolutional deep neural network to estimate a likelihood that a pixel in an image belongs to a foreground object. After receiving the image, the likelihood that the pixel in the image is part of the foreground object as opposed to background is then determined using the trained convolutional deep neural network.

Abstract: A method, system and computer program product for efficiently identifying images, videos, audio files or documents relevant to a user. Using either manual annotations or learned functions, the method predicts the relative strength of an attribute in an image, video, audio file or document from a pool of images, videos, audio files or documents. At query time, the system presents an initial set of reference images, videos, audio files or documents, and the user selects among them to provide relative attribute feedback. Using the resulting constraints in the multi-dimensional attribute space, the relevance function for the pool of images, videos, audio files or documents is updated and the relevance of the pool of images, videos, audio files or documents is re-computed. This procedure iterates using the accumulated constraints until the top-ranked images, videos, audio files or documents are acceptably close to the user's envisioned image, video, audio file or document.

Abstract: A method, system and computer program product for efficiently identifying images, videos, audio files or documents relevant to a user using binary search trees in attribute space for guiding relevance feedback. A binary tree is constructed for each relative attribute of interest. A “pivot exemplar” (at a node of the binary tree) is set for each relative attribute's binary tree as corresponding to the database image, video, audio file or document with a median relative attribute value among that subtree's child examples. A pivot exemplar out of the available current pivot exemplars that has the highest expected information gain is selected to be provided to the user. Comparative attribute feedback is then received from the user regarding whether a degree of the attribute in the user's target image, video, audio file or document is more, less or equal with the attribute displayed in the selected pivot exemplar.

Abstract: A method, system and computer program product for efficiently identifying images, videos, audio files or documents relevant to a user. Using either manual annotations or learned functions, the method predicts the relative strength of an attribute in an image, video, audio file or document from a pool of images, videos, audio files or documents. At query time, the system presents an initial set of reference images, videos, audio files or documents, and the user selects among them to provide relative attribute feedback. Using the resulting constraints in the multi-dimensional attribute space, the relevance function for the pool of images, videos, audio files or documents is updated and the relevance of the pool of images, videos, audio files or documents is re-computed. This procedure iterates using the accumulated constraints until the top-ranked images, videos, audio files or documents are acceptably close to the user's envisioned image, video, audio file or document.

Abstract: A method, system and computer program product for efficiently identifying images, videos, audio files or documents relevant to a user using binary search trees in attribute space for guiding relevance feedback. A binary tree is constructed for each relative attribute of interest. A “pivot exemplar” (at a node of the binary tree) is set for each relative attribute's binary tree as corresponding to the database image, video, audio file or document with a median relative attribute value among that subtree's child examples. A pivot exemplar out of the available current pivot exemplars that has the highest expected information gain is selected to be provided to the user. Comparative attribute feedback is then received from the user regarding whether a degree of the attribute in the user's target image, video, audio file or document is more, less or equal with the attribute displayed in the selected pivot exemplar.

Abstract: A method for classifying or comparing objects includes detecting points of interest within two objects, computing feature descriptors at said points of interest, forming a multi-resolution histogram over feature descriptors for each object and computing a weighted intersection of multi-resolution histogram for each object. An alternative embodiment includes a method for matching objects by defining a plurality of bins for multi-resolution histograms having various levels and a plurality of cluster groups, each group having a center, for each point of interest, calculating a bin index, a bin count and a maximal distance to the bin center and providing a path vector indicative of the bins chosen at each level. Still another embodiment includes a method for matching objects comprising creating a set of feature vectors for each object of interest, mapping each set of feature vectors to a single high-dimensional vector to create an embedding vector and encoding each embedding vector with a binary hash string.

Abstract: A method for classifying or comparing objects includes detecting points of interest within two objects, computing feature descriptors at said points of interest, forming a multi-resolution histogram over feature descriptors for each object and computing a weighted intersection of multi-resolution histogram for each object. An alternative embodiment includes a method for matching objects by defining a plurality of bins for multi-resolution histograms having various levels and a plurality of cluster groups, each group having a center, for each point of interest, calculating a bin index, a bin count and a maximal distance to the bin center and providing a path vector indicative of the bins chosen at each level. Still another embodiment includes a method for matching objects comprising creating a set of feature vectors for each object of interest, mapping each set of feature vectors to a single high-dimensional vector to create an embedding vector and encoding each embedding vector with a binary hash string.