There is an error in Eq. 13. denominator of Eq. 13 should be the magnitude of the first derivative of b(s) cubed (|db(s)/ds|^3) instead

Quantification of behaviour is essential for biology. Since the whisker system is a popular model, it is important to have methods for measuring whisker movements from behaving animals. Here, we developed a high-speed imaging system that measures whisker movements simultaneously from two vantage points. We developed a whisker tracker algorithm that automatically reconstructs 3D whisker information directly from the 'stereo' video data. The tracker is controlled via a Graphical User Interface that also allows user-friendly curation. The algorithm tracks whiskers, by fitting a 3D Bezier curve to the basal section of each target whisker. By using prior knowledge of natural whisker motion and natural whisker shape to constrain the fits and by minimising the number of fitted parameters, the algorithm is able to track multiple whiskers in parallel with low error rate. We used the output of the tracker to produce a 3D description of each tracked whisker, including its 3D orientation and 3D shape, as well as bending-related mechanical force. In conclusion, we present a non-invasive, automatic system to track whiskers in 3D from high-speed video, creating the opportunity for comprehensive 3D analysis of sensorimotor behaviour and its neural basis.

Motivated by the impact of superresolution methods for imaging, we undertake a detailed and systematic analysis of localization acuity for a biomimetic fingertip and a flat region of tactile skin. We identify three key factors underlying superresolution that enable the perceptual acuity to surpass the sensor resolution: 1) the sensor is constructed with multiple overlapping, broad but sensitive receptive fields; 2) the tactile perception method interpolates between receptors (taxels) to attain subtaxel acuity; and 3) active perception ensures robustness to unknown initial contact location. All factors follow from active Bayesian perception applied to biomimetic tactile sensors with an elastomeric covering that spreads the contact over multiple taxels. In consequence, we attain extreme superresolution with a 35-fold improvement of localization acuity (0.12 mm) over sensor resolution (4 mm). We envisage that these principles will enable cheap high-acuity tactile sensors that are highly customizable to suit their robotic use. Practical applications encompass any scenario where an end-effector must be placed accurately via the sense of touch.

Future robots may need to navigate where visual sensors fail. Touch sensors provide an alternative modality, largely unexplored in the context of robotic map building. We present the first results in grid based simultaneous localisation and mapping (SLAM) with biomimetic whisker sensors, and show how multi-whisker features coupled with priors about straight edges in the world can boost its performance. Our results are from a simple, small environment but are intended as a first baseline to measure future algorithms against.

Tactile sensing can be an important source of information for robots, and texture discrimination in particular is useful in object recognition and terrain identification. Whisker based tactile sensing has recently been shown to be a promising approach for mobile robots, using simple sensors and many classification approaches. However these approaches have often been tested in limited environments, and have not been compared against one another in a controlled way. A wide range of whisker-object contact poses are possible on a mobile robot, and the effect such contact variability has on sensing has not been properly investigated. We present a novel, carefully controlled study of simple surface texture classifiers on a large set of varied pose conditions that mimic those encountered by mobile robots. Namely, single brief whisker contacts with textured surfaces at a range of surface orientations and contact speeds. Results show that different classifiers are appropriate for different settings, with spectral template and feature based approaches performing best in surface texture, and contact speed estimation, respectively. The results may be used to inform selection of classifiers in tasks such as tactile SLAM.

Recent advances in modeling animal perception has motivated an approach of Bayesian perception applied to biomimetic robots. This study presents an initial application of Bayesian perception on an iCub fingertip sensor mounted on a dedicated positioning robot. We systematically probed the test system with five cylindrical stimuli offset by a range of positions relative to the fingertip. Testing the real-time speed and accuracy of shape and position discrimination, we achieved sub-millimeter accuracy with just a few taps. This result is apparently the first explicit demonstration of perceptual hyperacuity in robot touch, in that object positions are perceived more accurately than the taxel spacing. We also found substantial performance gains when the fingertip can reposition itself to avoid poor perceptual locations, which indicates that improved robot perception could mimic active perception in animals.

Many rodents use their whiskers to distinguish objects by surface texture. To examine possible mechanisms for this discrimination, data from an artificial whisker attached to a moving robot was used to test texture classification algorithms. This data was examined previously using a template-based classifier of the whisker vibration power spectrum. Motivated by a proposal about the neural computations underlying sensory decision making, we classified the raw whisker signal using the related `naive Bayes' method. The integration time window is important, with roughly 100ms of data required for good decisions and 500ms for the best decisions. For stereotyped motion, the classifier achieved hit rates of about 80% using a single (horizontal or vertical) stream of vibration data and 90% using both streams. Similar hit rates were achieved on natural data, apart from a single case in which the performance was only about 55%. Therefore this application of naive Bayes represents a biologically motivated algorithm that can perform well in a real-world robot task.