Search for: All records

Fitting facial landmarks on unconstrained videos is a challenging task with broad applications. Both generic and joint alignment methods have been proposed with varying degrees of success. However, many generic methods are heavily sensitive to initializations and usually rely on offline-trained static models, which limit their performance on sequential images with extensive variations. On the other hand, joint methods are restricted to offline applications, since they require all frames to conduct batch alignment. To address these limitations, we propose to exploit incremental learning for personalized ensemble alignment. We sample multiple initial shapes to achieve image congealing within one frame, which enables us to incrementally conduct ensemble alignment by group-sparse regularized rank minimization. At the same time, incremental subspace adaptation is performed to achieve personalized modeling in a unified framework. To alleviate the drifting issue, we leverage a very efficient fitting evaluation network to pick out well-aligned faces for robust incremental learning. Extensive experiments on both controlled and unconstrained datasets have validated our approach in different aspects and demonstrated its superior performance compared with state of the arts in terms of fitting accuracy and efficiency. 

We report on the high success rates of our new, scalable, computational approach for sign recognition from monocular video, exploiting linguistically annotated ASL datasets with multiple signers. We recognize signs using a hybrid framework combining state-of-the-art learning methods with features based on what is known about the linguistic composition of lexical signs. We model and recognize the sub-components of sign production, with attention to hand shape, orientation, location, motion trajectories, plus non-manual features, and we combine these within a CRF framework. The effect is to make the sign recognition problem robust, scalable, and feasible with relatively smaller datasets than are required for purely data-driven methods. From a 350-sign vocabulary of isolated, citation-form lexical signs from the American Sign Language Lexicon Video Dataset (ASLLVD), including both 1- and 2-handed signs, we achieve a top-1 accuracy of 93.3% and a top-5 accuracy of 97.9%. The high probability with which we can produce 5 sign candidates that contain the correct result opens the door to potential applications, as it is reasonable to provide a sign lookup functionality that offers the user 5 possible signs, in decreasing order of likelihood, with the user then asked to select the desired sign. 

2017 marked the release of a new version of SignStream® software, designed to facilitate linguistic analysis of ASL video. SignStream® provides an intuitive interface for labeling and time-aligning manual and non-manual components of the signing. Version 3 has many new features. For example, it enables representation of morpho-phonological information, including display of handshapes. An expanding ASL video corpus, annotated through use of SignStream®, is shared publicly on the Web. This corpus (video plus annotations) is Web-accessible—browsable, searchable, and downloadable—thanks to a new, improved version of our Data Access Interface: DAI 2. DAI 2 also offers Web access to a brand new Sign Bank, containing about 10,000 examples of about 3,000 distinct signs, as produced by up to 9 different ASL signers. This Sign Bank is also directly accessible from within SignStream®, thereby boosting the efficiency and consistency of annotation; new items can also be added to the Sign Bank. Soon to be integrated into SignStream® 3 and DAI 2 are visualizations of computer-generated analyses of the video: graphical display of eyebrow height, eye aperture, and head position. These resources are publicly available, for linguistic and computational research and for those who use or study AS 

We introduce a new general framework for sign recognition from monocular video using limited quantities of annotated data. The novelty of the hybrid framework we describe here is that we exploit state-of-the art learning methods while also incorporating features based on what we know about the linguistic composition of lexical signs. In particular, we analyze hand shape, orientation, location, and motion trajectories, and then use CRFs to combine this linguistically significant information for purposes of sign recognition. Our robust modeling and recognition of these sub-components of sign production allow an efficient parameterization of the sign recognition problem as compared with purely data-driven methods. This parameterization enables a scalable and extendable time-series learning approach that advances the state of the art in sign recognition, as shown by the results reported here for recognition of isolated, citation-form, lexical signs from American Sign Language (ASL).