Gestural communication of music structure during solo classical piano performance

The production and perception of music is a multimodal activity involving auditory, visual and conceptual processing, integrating these with prior knowledge and environmental experience. Musicians utilise expressive physical nuances to highlight salient features of the score. The question arises within the literature as to whether performers’ non-technical, non-sound-producing movements may be communicatively meaningful and convey important structural information to audience members and co-performers. In the light of previous performance research (Vines et al., 2006, Wanderley, 2002, Davidson, 1993), and considering findings within co-speech gestural research and auditory and audio-visual neuroscience, this thesis examines the nature of those movements not directly necessary for the production of sound, and their particular influence on audience perception. Within the current research 3D performance analysis is conducted using the Vicon 12- camera system and Nexus data-processing software. Performance gestures are identified as repeated patterns of motion relating to music structure, which not only express phrasing and structural hierarchy but are consistently and accurately interpreted as such by a perceiving audience. Gestural characteristics are analysed across performers and performance style using two Chopin preludes selected for their diverse yet comparable structures (Opus 28:7 and 6). Effects on perceptual judgements of presentation modes (visual-only, auditory-only, audiovisual, full- and point-light) and viewing conditions are explored. This thesis argues that while performance style is highly idiosyncratic, piano performers reliably generate structural gestures through repeated patterns of upper-body movement. The shapes and locations of phrasing motions are identified particular to the sample of performers investigated. Findings demonstrate that despite the personalised nature of the gestures, performers use increased velocity of movements to emphasise musical structure and that observers accurately and consistently locate phrasing junctures where these patterns and variation in motion magnitude, shape and velocity occur. By viewing performance motions in polar (spherical) rather than cartesian coordinate space it is possible to get mathematically closer to the movement generated by each of the nine performers, revealing distinct patterns of motion relating to phrasing structures, regardless of intended performance style. These patterns are highly individualised both to each performer and performed piece. Instantaneous velocity analysis indicates a right-directed bias of performance motion variation at salient structural features within individual performances. Perceptual analyses demonstrate that audience members are able to accurately and effectively detect phrasing structure from performance motion alone. This ability persists even for degraded point-light performances, where all extraneous environmental information has been removed. The relative contributions of audio, visual and audiovisual judgements demonstrate that the visual component of a performance does positively impact on the over- all accuracy of phrasing judgements, indicating that receivers are most effective in their recognition of structural segmentations when they can both see and hear a performance. Observers appear to make use of a rapid online judgement heuristics, adjusting response processes quickly to adapt and perform accurately across multiple modes of presentation and performance style. In line with existent theories within the literature, it is proposed that this processing ability may be related to cognitive and perceptual interpretation of syntax within gestural communication during social interaction and speech. Findings of this research may have future impact on performance pedagogy, computational analysis and performance research, as well as potentially influencing future investigations of the cognitive aspects of musical and gestural understanding.

The role of context in image annotation and recommendation

With the rise of smart phones, lifelogging devices (e.g. Google Glass) and popularity of image sharing websites (e.g. Flickr), users are capturing and sharing every aspect of their life online producing a wealth of visual content. Of these uploaded images, the majority are poorly annotated or exist in complete semantic isolation making the process of building retrieval systems difficult as one must firstly understand the meaning of an image in order to retrieve it. To alleviate this problem, many image sharing websites offer manual annotation tools which allow the user to “tag” their photos, however, these techniques are laborious and as a result have been poorly adopted; Sigurbjörnsson and van Zwol (2008) showed that 64% of images uploaded to Flickr are annotated with < 4 tags. Due to this, an entire body of research has focused on the automatic annotation of images (Hanbury, 2008; Smeulders et al., 2000; Zhang et al., 2012a) where one attempts to bridge the semantic gap between an image’s appearance and meaning e.g. the objects present. Despite two decades of research the semantic gap still largely exists and as a result automatic annotation models often offer unsatisfactory performance for industrial implementation. Further, these techniques can only annotate what they see, thus ignoring the “bigger picture” surrounding an image (e.g. its location, the event, the people present etc). Much work has therefore focused on building photo tag recommendation (PTR) methods which aid the user in the annotation process by suggesting tags related to those already present. These works have mainly focused on computing relationships between tags based on historical images e.g. that NY and timessquare co-exist in many images and are therefore highly correlated. However, tags are inherently noisy, sparse and ill-defined often resulting in poor PTR accuracy e.g. does NY refer to New York or New Year? This thesis proposes the exploitation of an image’s context which, unlike textual evidences, is always present, in order to alleviate this ambiguity in the tag recommendation process. Specifically we exploit the “what, who, where, when and how” of the image capture process in order to complement textual evidences in various photo tag recommendation and retrieval scenarios. In part II, we combine text, content-based (e.g. # of faces present) and contextual (e.g. day-of-the-week taken) signals for tag recommendation purposes, achieving up to a 75% improvement to precision@5 in comparison to a text-only TF-IDF baseline. We then consider external knowledge sources (i.e. Wikipedia & Twitter) as an alternative to (slower moving) Flickr in order to build recommendation models on, showing that similar accuracy could be achieved on these faster moving, yet entirely textual, datasets. In part II, we also highlight the merits of diversifying tag recommendation lists before discussing at length various problems with existing automatic image annotation and photo tag recommendation evaluation collections. In part III, we propose three new image retrieval scenarios, namely “visual event summarisation”, “image popularity prediction” and “lifelog summarisation”. In the first scenario, we attempt to produce a rank of relevant and diverse images for various news events by (i) removing irrelevant images such memes and visual duplicates (ii) before semantically clustering images based on the tweets in which they were originally posted. Using this approach, we were able to achieve over 50% precision for images in the top 5 ranks. In the second retrieval scenario, we show that by combining contextual and content-based features from images, we are able to predict if it will become “popular” (or not) with 74% accuracy, using an SVM classifier. Finally, in chapter 9 we employ blur detection and perceptual-hash clustering in order to remove noisy images from lifelogs, before combining visual and geo-temporal signals in order to capture a user’s “key moments” within their day. We believe that the results of this thesis show an important step towards building effective image retrieval models when there lacks sufficient textual content (i.e. a cold start).

Effects of modality, urgency and situation on responses to multimodal warnings for drivers

Signifying road-related events with warnings can be highly beneficial, especially when imminent attention is needed. This thesis describes how modality, urgency and situation can influence driver responses to multimodal displays used as warnings. These displays utilise all combinations of audio, visual and tactile modalities, reflecting different urgency levels. In this way, a new rich set of cues is designed, conveying information multimodally, to enhance reactions during driving, which is a highly visual task. The importance of the signified events to driving is reflected in the warnings, and safety-critical or non-critical situations are communicated through the cues. Novel warning designs are considered, using both abstract displays, with no semantic association to the signified event, and language-based ones, using speech. These two cue designs are compared, to discover their strengths and weaknesses as car alerts. The situations in which the new cues are delivered are varied, by simulating both critical and non-critical events and both manual and autonomous car scenarios. A novel set of guidelines for using multimodal driver displays is finally provided, considering the modalities utilised, the urgency signified, and the situation simulated.