Hidden Conditional Random Fields for Visual Speech Recognition

In this paper we present the application of Hidden Conditional Random Fields (HCRFs) to modelling speech for visual speech recognition. HCRFs may be easily adapted to model long range dependencies across an observation sequence. As a result visual word recognition performance can be improved as the model is able to take more of a contextual approach to generating state sequences. Results are presented from a speaker-dependent, isolated digit, visual speech recognition task using comparisons with a baseline HMM system. We firstly illustrate that word recognition rates on clean video using HCRFs can be improved by increasing the number of past and future observations being taken into account by each state. Secondly we compare model performances using various levels of video compression on the test set. As far as we are aware this is the first attempted use of HCRFs for visual speech recognition.

Intelligent Assembly Time Analysis Using a Digital Knowledge-Based Approach

The implementation of effective time analysis methods fast and accurately in the era of digital manufacturing has become a significant challenge for aerospace manufacturers hoping to build and maintain a competitive advantage. This paper proposes a structure oriented, knowledge-based approach for intelligent time analysis of aircraft assembly processes within a digital manufacturing framework. A knowledge system is developed so that the design knowledge can be intelligently retrieved for implementing assembly time analysis automatically. A time estimation method based on MOST, is reviewed and employed. Knowledge capture, transfer and storage within the digital manufacturing environment are extensively discussed. Configured plantypes, GUIs and functional modules are designed and developed for the automated time analysis. An exemplar study using an aircraft panel assembly from a regional jet is also presented. Although the method currently focuses on aircraft assembly, it can also be well utilized in other industry sectors, such as transportation, automobile and shipbuilding. The main contribution of the work is to present a methodology that facilitates the integration of time analysis with design and manufacturing using a digital manufacturing platform solution.

A study of the biological effects of modulated 6 MV radiation fields

The delivery of spatially modulated radiation fields has been shown to impact on in vitro cell survival responses. To study the effect of modulated fields on cell survival, dose response curves were determined for human DU-145 prostate, T98G glioma tumour cells and normal primary AGO-1552 fibroblast cells exposed to modulated and non-modulated field configurations delivered using a 6 MV Linac with multi-leaf collimator. When exposed to uniform fields delivered as a non-modulated or modulated configuration, no significant differences in survival were observed with the exception of DU-145 cells at a dose of 8 Gy (p = 0.024). Survival responses were determined for exposure to non-uniform-modulated beams in DU-145 and T98G and showed no deviation from the survival response observed following uniform non-modulated exposures. The results of these experiments indicate no major deviation in response to modulated fields compared to uniform exposures.