Deep Head Pose: Gaze-Direction Estimation in Multimodal Video

In this paper we present a convolutional neuralnetwork (CNN)-based model for human head pose estimation inlow-resolution multi-modal RGB-D data. We pose the problemas one of classification of human gazing direction. We furtherfine-tune a regressor based on the learned deep classifier. Next wecombine the two models (classification and regression) to estimateapproximate regression confidence. We present state-of-the-artresults in datasets that span the range of high-resolution humanrobot interaction (close up faces plus depth information) data tochallenging low resolution outdoor surveillance data. We buildupon our robust head-pose estimation and further introduce anew visual attention model to recover interaction with theenvironment. Using this probabilistic model, we show thatmany higher level scene understanding like human-human/sceneinteraction detection can be achieved. Our solution runs inreal-time on commercial hardware

An adaptive motion model for person tracking with instantaneous head-pose features

This letter presents novel behaviour-based tracking of people in low-resolution using instantaneous priors mediated by head-pose. We extend the Kalman Filter to adaptively combine motion information with an instantaneous prior belief about where the person will go based on where they are currently looking. We apply this new method to pedestrian surveillance, using automatically-derived head pose estimates, although the theory is not limited to head-pose priors. We perform a statistical analysis of pedestrian gazing behaviour and demonstrate tracking performance on a set of simulated and real pedestrian observations. We show that by using instantaneous `intentional' priors our algorithm significantly outperforms a standard Kalman Filter on comprehensive test data.

LOTUS: A low cost, ultraviolet spectrograph

We describe the design, construction and commissioning of LOTUS; a simple, low-cost long-slit spectrograph for the Liverpool Telescope. The design is optimized for near-UV and visible wavelengths and uses all transmitting optics. It exploits the instrument focal plane field curvature to partially correct axial chromatic aberration. A stepped slit provides narrow (2.5x95 arcsec) and wide (5x25 arcsec) options that are optimized for spectral resolution and flux calibration respectively. On sky testing shows a wavelength range of 3200-6300 Angstroms with a peak system throughput (including detector quantum efficiency) of 15 per cent and wavelength dependant spectral resolution of R=225-430. By repeated observations of the symbiotic emission line star AG Peg we demonstrate the wavelength stability of the system is less than 2 Angstroms rms and is limited by the positioning of the object in the slit. The spectrograph is now in routine operation monitoring the activity of comet 67P/Churyumov-Gerasimenko during its current post-perihelion apparition.

High resolution Thomson Parabola Spectrometer for full spectral capture of multi-species ion beams

We report on the experimental characterisation of laser-driven ion beams using a Thomson Parabola Spectrometer (TPS) equipped with trapezoidally shaped electric plates, proposed by Gwynne et al. [Rev. Sci. Instrum. 85, 033304 (2014)]. While a pair of extended (30 cm long) electric plates was able to produce a significant increase in the separation between neighbouring ion species at high energies, deploying a trapezoidal design circumvented the spectral clipping at the low energy end of the ion spectra. The shape of the electric plate was chosen carefully considering, for the given spectrometer configuration, the range of detectable ion energies and species. Analytical tracing of the ion parabolas matches closely with the experimental data, which suggests a minimal effect of fringe fields on the escaping ions close to the wedged edge of the electrode. The analytical formulae were derived considering the relativistic correction required for the high energy ions to be characterised using such spectrometer.