Abstract: Aspects of this disclosure include technologies for detecting mislabeled products. In one embodiment, the disclosed system will capture an image of a product when the MRL of the product is scanned or being scanned. After recognizing the product in the image, the size of the area containing the product may be calculated. Subsequently, the disclosed system can determine whether the MRL mismatches the product in the image if this size of the area containing the product does not match the standard size associated with the MRL.

Abstract: Weakly supervised instance segmentation refers to the task of training a system to detect object locations and segment instances of the detected objects, where the training data includes only images and image-level labels. This disclosure includes an enhanced pipeline and enhanced training methods that progressively mine pixel-wise labels, when trained via image-level labels. Four cascaded modules are employed, including: a multi-label classification module, an object detection module, an instance refinement module, and instance segmentation module. The modules share a common backbone. The cascaded pipeline is trained alternatively with a curriculum learning strategy which generalizes image level supervision to pixel level supervision, and a post validation training stage, which runs in the inverse order. In the curriculum learning stage, a proposal refinement sub-module is employed to locate object parts and finding key pixels during classification.

Abstract: Aspects of this disclosure include technologies to detect unpackaged, unlabeled, or mislabeled products based on product images. Leveraging from improved machine learning techniques, the disclosed technical solution can reduce a full product space for product search to a partial product space. Accordingly, a limited number of product candidates in the partial product space may be visually presented for user evaluation. Sometimes, a product candidate is to be comparatively presented with a live image of the product via a uniquely designed graphic user interface, which further improves the confidence of the user and the accuracy of the underlying transaction.

Abstract: Methods and systems are provided for generating a multi-label classification system. The multi-label classification system can use a multi-label classification neural network system to identify one or more labels for an image. The multi-label classification system can explicitly take into account the relationship between classes in identifying labels. A relevance sub-network of the multi-label classification neural network system can capture relevance information between the classes. Such a relevance sub-network can decouple independence between classes to focus learning on relevance between the classes.

Abstract: Aspects of this disclosure include technologies to detect unpackaged, unlabeled, or mislabeled products based on product images. Leveraging from improved machine learning techniques, the disclosed technical solution can reduce a full product space for product search to a partial product space. Accordingly, a limited number of product candidates in the partial product space may be visually presented for user evaluation. Sometimes, a product candidate is to be comparatively presented with a live image of the product via a uniquely designed graphic user interface, which further improves the confidence of the user and the accuracy of the underlying transaction.

Abstract: Aspects of this disclosure include technologies for detecting mislabeled products. In one embodiment, the disclosed system will capture an image of a product when the MRL of the product is scanned or being scanned. After recognizing the product in the image, the size of the area containing the product may be calculated. Subsequently, the disclosed system can determine whether the MRL mismatches the product in the image if this size of the area containing the product does not match the standard size associated with the MRL.

Abstract: An oral care product array includes a handle component and multiple refill components therefor. The handle component includes a power source. A first refill component includes a first light generating module which emits light in a first spectrum from the first refill component. A second refill component includes a second light generating module which emits light in a second spectrum from the second refill component. Each refill component has a coupled state, when coupled to the handle component, and a decoupled state, when decoupled from the handle component. When the first refill component is in the coupled state, the first light generating module is powered by the power source to emit light in the first spectrum. When the second refill component is in the coupled state, the second light generating module is powered by the power source to emit light in the second spectrum.

Abstract: The present disclosure provides methods of conducting and analyzing mouse clinical trials. In one embodiment, the method comprises the steps of receiving a dataset of tumor volumes measured in a mouse clinical trial, determining tumor growth curve of the treatment group and tumor growth curve of the control group; determining area under curve (AUC) of the treatment group and AUC of the control group; and evaluating efficacy of the drug based on an AUC ratio between the AUC of the treatment group and the AUC of the control group, wherein the mouse clinical trial comprises the steps of: obtaining a tumor sample derived from a patient; grafting the tumor sample to a treatment group comprising m mice and a control group comprising n mice, wherein m and n are integers; treating the treatment group with a drug; treating the control group with a vehicle; and measuring tumor volume of the treatment group and tumor volume of the control group at a plurality of days.

Abstract: Aspects of this disclosure include technologies for detecting mislabeled products. In one embodiment, the disclosed system will capture an image of a product when the MRL of the product is scanned or being scanned. After recognizing the product in the image, the size of the area containing the product may be calculated. Subsequently, the disclosed system can determine whether the MRL mismatches the product in the image if this size of the area containing the product does not match the standard size associated with the MRL.

Abstract: This disclosure includes technologies for video recognition in general. The disclosed system can automatically detect various types of actions in a video, including reportable actions that cause shrinkage in a practical application for loss prevention in the retail industry. The temporal evolution of spatio-temporal features in the video are used for action recognition. Such features may be learned via a 4D convolutional operation, which is adapted to model low-level features based on a residual 4D block. Further, appropriate responses may be invoked if a reportable action is recognized.

Abstract: An optical plate includes at least two first optical portions for respectively covering two light emitting modules, and at least one second optical portion for covering a gap between the light emitting modules. Each first optical portion receives light rays emitted from a respective light emitting module and causes the same to emit along a predetermined direction. The second optical portion has first and second inner inclined surfaces, and first and second outer inclined surfaces. The inner inclined surfaces guide the light rays emitted from the respective light emitting modules toward the outer inclined surfaces, respectively, such that the light rays emitting out therefrom are parallel to the predetermined direction.

Abstract: An oral care product array includes a handle component and multiple refill components therefor. The handle component includes a power source. A first refill component includes a first light generating module which emits light in a first spectrum from the first refill component. A second refill component includes a second light generating module which emits light in a second spectrum from the second refill component. Each refill component has a coupled state, when coupled to the handle component, and a decoupled state, when decoupled from the handle component. When the first refill component is in the coupled state, the first light generating module is powered by the power source to emit light in the first spectrum. When the second refill component is in the coupled state, the second light generating module is powered by the power source to emit light in the second spectrum.

Abstract: A light-emitting oral care implement includes a light delivery unit that includes a passageway through a body of the oral care implement and a tuft comprising a plurality of filaments located in the passageway, each of the filaments having a core component and a sheath component. The core component may have a transmittance that is greater than the transmittance of the sheath component. The core component may include a base portion that protrudes from a lower end of the sheath component, the base portion forming an anchor portion of the tuft that anchors the tuft to the body of the oral care implement. Furthermore, the oral care implement may include a light source that transmits lights through the anchor portion of the tuft and through the core components for transmittance from the filaments at a location above a front surface of the body.

Abstract: A monofilament bristle (200) is disclosed that comprises coextruded core (210) and sheath components (220). The sheath component (220) has an outer surface (218) and surrounds a first portion (215) of the core component (210). A second portion (216) of the core component (210) protrudes from a top terminal end (203) of the sheath component (220) to form an exposed tip portion (203). In one embodiment, the exposed tip portion (203) of the core component (210) has an outer surface (217) that includes friction-enhancing features. In another embodiment, the exposed tip portion (203) is tapered and has an outer surface (217) with a first roughness and the outer surface (218) of the sheath component (220) has a second roughness, the second roughness being greater than the first roughness. In a further embodiment, the core component (210) has an outer surface (217) configured to have higher coefficient of friction than the outer surface (218) of the sheath component (220).

Abstract: Embodiments of the present invention provide systems, methods, and computer storage media for 3D segmentation. A 3D segmentation network can perform a voxel-wise classification of a 3D volume such as a brain tumor. The 3D segmentation network can accept a plurality of 3D representations of the 3D volume (e.g., MRI modalities) into corresponding 3D input channels. Generally, 3D convolutions can be applied by convolutional layers of the 3D segmentation network. Convolutional blocks can generate successive resolutions of feature maps from the plurality of 3D representations by downsampling outputs from prior convolutional blocks. The feature maps can be upsampled and combined to aggregate local and global detail from multiple resolutions. A multi-class classifier can be applied to each voxel at the output layer to generate a voxel-wise prediction map with the same spatial size as the inputs.

Abstract: Embodiments of the present invention provide systems, methods, and computer storage media for 3D segmentation. A 3D segmentation network can perform a voxel-wise classification of a 3D volume such as a brain tumor. The 3D segmentation network can accept a plurality of 3D representations of the 3D volume (e.g., MRI modalities) into corresponding 3D input channels. Generally, 3D convolutions can be applied by convolutional layers of the 3D segmentation network. Convolutional blocks can generate successive resolutions of feature maps from the plurality of 3D representations by downsampling outputs from prior convolutional blocks. The feature maps can be upsampled and combined to aggregate local and global detail from multiple resolutions. A multi-class classifier can be applied to each voxel at the output layer to generate a voxel-wise prediction map with the same spatial size as the inputs.

Abstract: A method for predicting the responsiveness of a cancer cell to an MEK inhibitor, comprising detecting the presence of at least one mutation in one or more genes selected from the group consisting of ADAM12, COL14A1, TNN, and TP53, in the cancer cell, by contacting a nucleic acid sample derived from the cancer cell with at least one oligonucleotide which allows specific detection of the mutation; wherein presence of mutation in ADAM12, COL14A1, TNN, TP53 and/or any combination thereof is indicative of decreased responsiveness of the cancer cell to the ERK inhibitor.

Abstract: An oral care implement including an elongated body having a head portion supporting plurality of tooth cleaning elements and a handle portion defining an outer surface. A grip control component is disposed on the handle portion and includes a pair of transversely spaced apart control surfaces separated by a longitudinally extending channel. The control surfaces are elevated above the outer surface of the handle portion and positioned to engage a user's thumb. The control component is operable to tilt the head portion at an upward or downward angle relative to a vertical plane of a user's teeth when the control component is grasped between the user's thumb and forefinger for producing a proper brushing angle at the gum line.

Abstract: A panel of gene expression markers for gastric cancer patient treated or to be treated by cetuximab is provided. Methods and compositions are also provided, e.g., kits, for evaluating gene expression levels of the markers and methods of using such gene expression levels to predict a gastric cancer patient's response to cetuximab.