Abstract: A method and a system for garment try-on. The method includes: capturing a three dimensional (3D) image of a customer; obtaining first 3D pose of the customer in the 3D image; performing a machine learning model on the 3D pose to generate a first skinned multi-person linear model (SMPL) pose; calculating an angle of the whole body rotation of the customer based on the first SMPL pose; when the angle is in a predefined range relative to a front direction of the customer: constructing an SMPL model using the first SMPL pose; and when the angle is out of the predefined range: generating a second SMPL pose using two dimensional (2D) component of the 3D image, and constructing the SMPL model using the second SMPL pose.

Abstract: A method and a system for recommending a target garment matching an inputted garment. The method includes: extracting attributes from text description and image of the inputted garment to obtain extracted attributes; querying a knowledge graph using the extracted attributes to obtain matched attributes; retrieving candidate products from a garment pool using the matched attributes; extracting features from the inputted garment and the candidate products; determining the target garment from the candidate products based on grading scores between the features of the inputted garment and the features of the candidate products; and recommending the target garment. The knowledge graph includes nodes corresponding to type of clothes, category of clothes, attribute keys, values of attribute keys, context keys, values of context keys, combination of the values of the attribute keys and the type of clothes, and combination of the value of the attribute keys and the category of clothes.

Abstract: A system for knowledge graph construction. The system includes a computing device. The computing device has a processor and a storage device storing computer executable code. The computer executable code, when executed at the processor, is configured to: define entities and relations of the knowledge graph; provide documents having sentences; convert the sentences into fix length sentence embeddings and regard the sentence embeddings as primary capsule layers; use a set transformer to learn entity capsules and relation capsules from the primary capsule layers; for each triple, project head and tail entities from entity space to the specific relation space, and determine the relation exists when the sum of the projected head entity vector and the relation vector substantially equals to the projected tail entity vector; and construct the knowledge graph using the head entity, the tail entity, and the determined relation.

Abstract: A system and method for autonomous navigation. The system includes a computing device having a processor and a storage device storing computer executable code. The computer executable code, when executed at the processor, is configured to: provide a planned path having intersections in an environment, where the intersections and roads therebetween are represented by sequential place identifications (IDs); receive images of the environment; perform convolutional neural network on the images to obtain predicted place IDs; when a predicted place ID of a current image is next to a place ID of a previous image, and is the same as predicted place IDs of a predetermined number of following images, define the predicted place ID as place IDs of the current and the following images; and perform autonomous navigation based on the planned path and the image place IDs.

Abstract: Systems and method for lifelong tag learning. The system includes a computing device having a processor and a storage device storing computer executable code. The computer executable code is configured to: provide product descriptions and seed tags characterizing products; train a named-entity recognition (NER) model using the product descriptions and the seed tags; predict pseudo tags from the product descriptions using the NER model; calculate confidence scores of the pseudo tags; compare the confidence scores with a threshold, and define the pseudo tags as true tags when the confidence scores are greater than the threshold; add the true tags to the seed tags to obtain updated tags; and repeat the steps of training, predicting, calculating, comparing and adding using the product descriptions and the updated tags, so as to keep updating the updated tags.

Abstract: A system and a method for training an inference model using a computing device. The method includes: providing a text-to-vector converter; providing the inference model and pre-training the inference model using labeled fashion entries; providing non-labeled fashion entries; separating each of the non-labeled fashion entries into a target image and target text; converting the target text into a category vector and an attribute vector using the text-to-vector converter; processing the target image using the inference model to obtain processed target image and target image label; comparing the category vector to the target image label; when the category vector matches the target image label, updating the target image label based on the category vector and the attribute vector to obtain updated label; and retraining the inference model using the processed target image and the updated label.

Abstract: System and method for relation extraction using adaptive thresholding and localized context pooling (ATLOP). The system includes a computing device, the computing device has a processer and a storage device storing computer executable code. The computer executable code is configured to provide a document; embed entities in the document into embedding vectors; and predict relations between a pair of entities in the document using their embedding vectors. The relation prediction is performed based on an improved language model. Each relation has an adaptive threshold, and the relation between the pair of entities is determined to exist when a logit of the relation between the pair of entities is greater than a logit function of the corresponding adaptive threshold.

Abstract: A system and a method for time series forecasting. The method includes: providing input feature vectors corresponding to a plurality of future time steps; performing bi-directional long-short term memory network (BiLSTM) on the input feature vectors to obtain hidden outputs corresponding to the plurality of future time steps; for each future time step: performing temporal convolution on the hidden outputs using a plurality of temporal scales to obtain context features at the plurality of temporal scales, and summating the context features at the plurality of temporal scales using a plurality of weights to obtain multi-scale context features; and converting the multi-scale context features to obtain the time series forecasting corresponding to the future time steps.

Abstract: A method and system for content display. The method includes: capturing, by an imaging device, a plurality of images of an environment; generating, by a computing device, first user profiles based on appearance of users recognized in the images; detecting devices in the environment and associating the detected devices to the users to obtain associated devices that are held by the users; retrieving, by the computing device, second user profiles of the users based on identifications of the associated devices; selecting at least one content based on the first user profiles and the second user profiles; and displaying the at least one selected content.

Abstract: A system for assessing text content of a product. The system includes a computing device having a processor and a storage device storing computer executable code. The computer executable code, when executed at the processor, is configured to: provide text contents and confounding features of products; train a first regression model using the text content and the confounding features of the products; train the second regression model using the confounding features; operate the first regression model using the text contents and the confounding features to obtain a total loss; operate the second regression model using the confounding features of to obtain a partial loss; subtract the total loss from the partial loss to obtain a residual loss; use the residual loss to evaluate models and parameters for the regression models; and use the first regression model to obtain log odds of the words indicating importance of the words.

Abstract: System and method multitask prediction. The system include a computing device. The computing device has a processer and a storage device storing computer executable code. The computer executable code is configured to: provide a head entity and a document containing the head entity; process the head entity and the document by a language model to obtain head extraction corresponding to the head entity, tail extractions corresponding to tail entities in the document, and sentence extraction corresponding to sentences in the document; predict a head-tail relation between the head extraction and the tail extractions using a first bilinear layer; combine the sentence extraction and a relation vector corresponding to the predicted head-tail relation using a second bilinear layer to obtain a sentence-relation combination; and predict an evidence sentence supporting the head-tail relation using a third bilinear layer based on the sentence-relation combination and attention extracted from the language model.

Abstract: A system and a method for obtaining a content plan are provided. The system includes a computing device configured to: provide a tag for each clause of each sentence of each product description of each of a category of products, where the tags include introductive tags corresponding to attributions of the products and description tags corresponding to descriptions of the products; determine tag exclusiveness, tag similarity, tag distance, and tag causality between the tags; generate candidate tag sequences the product descriptions; and filter the candidate tag sequences using the tag exclusiveness, the tag similarity, the tag distance and the tag causality to obtain the content plan.

Abstract: A computer-implemented method, a device, and a non-transitory computer-readable storage medium of automatically determining an interactive GUI element in a graphic user interface (GUI) to be interacted. The method includes: detecting, by the processor, one or more candidate interactive GUI elements in the GUI based on a plurality of algorithms; determining, by the processor, a likelihood indicator for each of the one or more candidate interactive GUI elements, a likelihood indicator indicating the likelihood that a candidate interactive GUI element associated with the likelihood indicator is an interactive GUI element to be interacted; and determining, by the processor, an interactive GUI element to be interacted from the one or more candidate interactive GUI elements based on the likelihood indicators.

Abstract: A method and system for intelligent guided shopping. The system includes a computing device. The computing device has a processor and a storage device storing computer executable code. The computer executable code, when executed at the processor, is configured to provide inference models based on user search data. Each of the inference models includes nodes and edges between the nodes. Each of the edges pointing from a first node to a second node is characterized with a probability value. The probability value represents a probability of searching a second keyword corresponding to the second node after searching a first keyword corresponding to the first node.

Abstract: System and method for simultaneous object detection and semantic segmentation. The system includes a computing device. The computing device has a processor and a non-volatile memory storing computer executable code. The computer executable code, when executed at the processor, is configured to: receive an image of a scene; process the image using a neural network backbone to obtain a feature map; process the feature map using an object detection module to obtain object detection result of the image; and process the feature map using a semantic segmentation module to obtain semantic segmentation result of the image. The object detection module and the semantic segmentation module are trained using a same loss function comprising an object detection component and a semantic segmentation component.

Abstract: A method, a device, and a non-transitory computer readable medium for item recommendation based on visual elements. The method includes: determining, by one or more processors, visual elements from an item image of an item; generating, by the one or more processors, an element descriptor for the item based on at least a part of the visual elements; and calculating, by the one or more processors, a compatibility value between the element descriptor and one or more other element descriptors for one or more other items.

Abstract: A system and a method for multimedia data analysis. The system includes a computing device. The computer device includes a processor and a storage device storing computer executable code. The computer executable code, when executed at the processor, is configured to: encode task description and task meta info into concatenated task vectors; encode context text and context meta info into concatenated context text vectors; encode context image and the context meta info into concatenated context image vectors; perform dual coattention on the concatenated task vectors and the concatenated context text and image vectors to obtain attended task vectors and attended context vectors; perform BiLSTM on the attended task vectors and the attended context vectors to obtain task encoding and context encoding; and decode the task encoding and the context encoding to obtain an answer to the task.

Abstract: A system for three dimensional (3D) model texturing. The system includes a computing device configured to: instruct a camera to capture images of an object; construct a 3D model of the object based on the images; texture the 3D model using keyframes of the images to obtain a textured 3D model; generate guidance images by projecting the textured 3D model to camera poses of the keyframes; for each guidance image and keyframe pair: modify the guidance image and the keyframe based on their correspondence to obtain enhanced guidance image and warped image, combine the enhanced guidance image and the warped image to obtain harmonized image, and project color from the harmonized images to the 3D model to obtain textured 3D model.

Abstract: A method for virtual fitting, includes providing a customer avatar and a garment model of a selected garment; registering the garment model to the customer avatar to obtain a dressed garment; capturing a motion image of the customer; performing simulation to the customer avatar and the dressed garment based on the motion image; rendering the simulated customer avatar-garment; and displaying the rendered avatar-garment on a displaying device. Rigging is performed to the dressed garment to obtain rigged garment, and cloth simulation is performed to the dressed garment model to obtain simulated cloth. Each vertex of the simulated cloth has a distance to nearest body mesh vertices on the simulated customer avatar, and the simulated garment is the rigged garment when the distance is less than 0 or the distance is greater than a threshold distance. Air mesh between the avatar and the garment is used for collision detection.

Abstract: A method and system for generating a three-dimensional (3D) model of a garment. The method includes: providing an image of the garment; generating a 3D garment representation based on the image; registering the 3D garment representation to a 3D model to obtain a preliminary 3D garment model; projecting the preliminary 3D garment model dressed on the 3D model to an 2D projected image; and comparing the 2D projected image with the image of the garment, so as to refine the preliminary 3D garment model to obtain the final 3D model of the garment.