Carnegie Mellon University

Khoa Luu

Marios Savvides

Chenchen Zhu

Yutong Zheng

Chandrasekhar Bhagavatula

Keshav Thirumalai Seshadri

Marios Sawides

The present invention provides a novel approach to simultaneously extracting the 3D shape of the face and the semantically consistent 2D alignment using a 3D Spatial Transformer Network (3DSTN) to model both the camera projection matrix and the warping parameters of a 3D model. By utilizing a generic 3D model and a thin plate spline (TPS) warping function, subject-specific 3D shapes are able to be generated without the need for a large 3D shape basis.

3D spatial transformer network

Methods of detecting an object in an image using a convolutional neural-network-based architecture that processes multiple feature maps of differing scales from differing convolution layers within a convolutional network to create a regional-proposal bounding box. The bounding box is projected back to the feature maps of the individual convolution layers to obtain a set of regions of interest (ROIs) and a corresponding set of context regions that provide additional context for the ROIs. These ROIs and context regions are processed to create a confidence score representing a confidence that the object detected in the bounding box is the desired object. These processes allow the method to utilize deep features encoded in both the global and the local representation for object regions, allowing the method to robustly deal with challenges in the problem of object detection. Software for executing the disclosed methods within an object-detection system is also disclosed.

Methods and software for detecting objects in an image using a contextual multiscale fast region-based convolutional neural network

Methods of detecting an object in an image using a convolutional neural network based architecture that processes multiple feature maps of differing scales from differing convolution layers within a convolutional network to create a regional-proposal bounding box. The bounding box is projected back to the feature maps of the individual convolution layers to obtain a set of regions of interest. These regions of interest are then processed to ultimately create a confidence score representing the confidence that the object detected in the bounding box is the desired object. These processes allow the method to utilize deep features encoded in both the global and the local representation for object regions, allowing the method to robustly deal with challenges in the problem of robust object detection. Software for executing the disclosed methods within an object-detection system is also disclosed.

Methods and software for detecting objects in images using a multiscale fast region-based convolutional neural network

Age-estimation of a face of an individual is represented in image data. In one embodiment, age-estimation techniques involves combining a Contourlet Appearance Model (CAM) for facial-age feature extraction and Support Vector Regression (SVR) for learning aging rules in order to improve the accuracy of age-estimation over the current techniques. In a particular example, characteristics of input facial images are converted to feature vectors by CAM, then these feature vectors are analyzed by an aging-mechanism-based classifier to estimate whether the images represent faces of younger or older people prior to age-estimation, the aging-mechanism-based classifier being generated in one embodiment by running Support Vector Machines (SVM) on training images. In an exemplary binary youth/adult classifier, faces classified as adults are passed to an adult age-estimation function and the others are passed to a youth age-estimation function.

Growing community of inventors

Pittsburgh, PA, United States of America

Khoa Luu