Speaker diarization : "who spoke when"

Speaker diarization is the process of annotating an input audio with information that attributes temporal regions of the audio signal to their respective sources, which may include both speech and non-speech events. For speech regions, the diarization system also specifies the locations of speaker boundaries and assign relative speaker labels to each homogeneous segment of speech. In short, speaker diarization systems effectively answer the question of ‘who spoke when’. There are several important applications for speaker diarization technology, such as facilitating speaker indexing systems to allow users to directly access the relevant segments of interest within a given audio, and assisting with other downstream processes such as summarizing and parsing. When combined with automatic speech recognition (ASR) systems, the metadata extracted from a speaker diarization system can provide complementary information for ASR transcripts including the location of speaker turns and relative speaker segment labels, making the transcripts more readable. Speaker diarization output can also be used to localize the instances of specific speakers to pool data for model adaptation, which in turn boosts transcription accuracies. Speaker diarization therefore plays an important role as a preliminary step in automatic transcription of audio data. The aim of this work is to improve the usefulness and practicality of speaker diarization technology, through the reduction of diarization error rates. In particular, this research is focused on the segmentation and clustering stages within a diarization system. Although particular emphasis is placed on the broadcast news audio domain and systems developed throughout this work are also trained and tested on broadcast news data, the techniques proposed in this dissertation are also applicable to other domains including telephone conversations and meetings audio. Three main research themes were pursued: heuristic rules for speaker segmentation, modelling uncertainty in speaker model estimates, and modelling uncertainty in eigenvoice speaker modelling. The use of heuristic approaches for the speaker segmentation task was first investigated, with emphasis placed on minimizing missed boundary detections. A set of heuristic rules was proposed, to govern the detection and heuristic selection of candidate speaker segment boundaries. A second pass, using the same heuristic algorithm with a smaller window, was also proposed with the aim of improving detection of boundaries around short speaker segments. Compared to single threshold based methods, the proposed heuristic approach was shown to provide improved segmentation performance, leading to a reduction in the overall diarization error rate. Methods to model the uncertainty in speaker model estimates were developed, to address the difficulties associated with making segmentation and clustering decisions with limited data in the speaker segments. The Bayes factor, derived specifically for multivariate Gaussian speaker modelling, was introduced to account for the uncertainty of the speaker model estimates. The use of the Bayes factor also enabled the incorporation of prior information regarding the audio to aid segmentation and clustering decisions. The idea of modelling uncertainty in speaker model estimates was also extended to the eigenvoice speaker modelling framework for the speaker clustering task. Building on the application of Bayesian approaches to the speaker diarization problem, the proposed approach takes into account the uncertainty associated with the explicit estimation of the speaker factors. The proposed decision criteria, based on Bayesian theory, was shown to generally outperform their non- Bayesian counterparts.

Emerging methods for digital media research : an introduction

Now as in earlier periods of acute change in the media environment, new disciplinary articulations are producing new methods for media and communication research. At the same time, established media and communication studies meth- ods are being recombined, reconfigured, and remediated alongside their objects of study. This special issue of JOBEM seeks to explore the conceptual, political, and practical aspects of emerging methods for digital media research. It does so at the conjuncture of a number of important contemporary trends: the rise of a ‘‘third wave’’ of the Digital Humanities and the ‘‘computational turn’’ (Berry, 2011) associated with natively digital objects and the methods for studying them; the apparently ubiquitous Big Data paradigm—with its various manifestations across academia, business, and government — that brings with it a rapidly increasing interest in social media communication and online ‘‘behavior’’ from the ‘‘hard’’ sciences; along with the multisited, embodied, and emplaced nature of everyday digital media practice.

Review : "Networking : commercial television in Australia : a history" -- Nick Herd

Nick Herd begins his institutional history of Australian commercial television in the early 1890s, when an amateur inventor named Henry Sutton designed the ‘telephane’ with the intent of watching the Melbourne Cup in his home town of Ballarat. The ‘race that stops a nation’ was not broadcast live on television until 1960, but Sutton’s initiative indicates how closely sport and television were aligned in Australia even before the medium existed. The first licensed commercial stations to begin regular broadcasting went on air in Sydney and Melbourne shortly before the 1956 Melbourne Olympic Games, although Herd claims that this was ‘almost accidental’ rather than planned. (49) Only Melbourne viewers were able to see some events live, many via television sets in Ampol service stations following the company’s last minute sponsorship of coverage on Melbourne station GTV-9...

The democratisation of journalism practice for a politically informed citizenry : the capacities of new media to engage citizens in public life - a case study

This research investigates the extent to which the World Wide Web and the participatory news media culture have contributed to the democratisation of journalism since 1997. It examined the different ways in which public service and commercial news media models use digital platforms to fulfil their obligations as members of the Fourth Estate. The research found that the digital environment provides news organisations with greater scope for transparency, interactivity, collaboration and social networking compared to the traditional print and broadcast platforms.

Capturing the 'authentic voice' : challenges and opportunities for voice and self-representation in two ABC storytelling projects

In recent years there has been a noticeable move by various public institutions, such as public service broadcasters and community media organisations, to capture and disseminate the voices and viewpoints of ‘ordinary people’ through inviting them to share stories about their lives. One of the foremost objectives of many such projects is to provide under-represented individuals and groups with an opportunity to express and represent themselves; as such, the capture and broadcast of ‘authentic voices’ is a central value. This paper discusses the notion of ‘authentic voice’, and questions the framing role of public media organisations in storytelling projects that aim to provide individuals with space for self-expression and self-representation. It considers the ways in which tensions arise on multiple levels when individuals are asked to express and represent themselves within projects and spaces that are managed by institutions. This paper begins by discussing the challenges and opportunities that arise within storytelling projects that are facilitated by public institutions and community media arts organisations, and that aim to amplify the voices of “ordinary people” (Thumim, 2009). It examines ways in which ‘voice’ is facilitated, curated, broadcast and distributed within such projects, particularly questioning the ways in which project facilitation and the curation of stories for public broadcast can both help and hinder the amplification of ‘authentic voice’. Furthermore, we seek to discuss how ‘authentic voice’ is defined, and what is involved in the process of amplification. The paper moves on to discuss a case study in order to demonstrate some of the tensions that are evident within a storytelling project that is managed by a public institution – Australia’s national broadcaster – and the ways these tensions impact upon the capture and broadcast of an ‘authentic voice’ for project participants. The Australian Broadcasting Corporation’s (ABC) ‘Heywire’ project is a storytelling competition and website that aims to ‘give voice’ to 16-22 year olds who live in rural, regional and remote parts of Australia. Looking at tensions that exist on organisational, political and philosophical levels within the Heywire project reveals a number of conflicts of interest and objectives between the institution and project participants. This leads us to question whether institutionally-managed storytelling projects can effectively support individuals to have an ‘authentic voice’, and whether struggles of aims and objectives diminish the personal benefits that people may derive from expressing and representing themselves within such projects.

Community media 2.0 : a report from the 'Co-Creative Communities' forum

Participatory digital culture presents major challenges to all traditional media outlets, but it presents very direct challenges to the community broadcast sector, which was established from the outset as local, community-driven and participatory. These and other issues were the focus of a recent forum at the Australian Centre for the Moving Image in Melbourne (Co-Creative Communities, 8–9 November 2012). The forum was part of a national research project, which has been exploring how Australian community arts and media organisations are responding to participatory digital culture, social media and user-led innovation. Focusing on the organisations who presented at the symposium, the paper examines how community-interest media is making the most of new and social media platforms. It considers examples of participatory digital media that have emerged from the community broadcast sector, but it also considers local, collaborative, community-interest media projects developed by public broadcasters and organisations involved in arts, social justice and development. Drawing on forum transcripts and follow-up research the essay describes some of the key trends shaping how community-interest media organisations and independent producers are working with participatory digital culture, and with what success.

A novel real time pricing scheme for demand response in residential distribution systems

Price based technique is one way to handle increase in peak demand and deal with voltage violations in residential distribution systems. This paper proposes an improved real time pricing scheme for residential customers with demand response option. Smart meters and in-home display units are used to broadcast the price and appropriate load adjustment signals. Customers are given an opportunity to respond to the signals and adjust the loads. This scheme helps distribution companies to deal with overloading problems and voltage issues in a more efficient way. Also, variations in wholesale electricity prices are passed on to electricity customers to take collective measure to reduce network peak demand. It is ensured that both customers and utility are benefitted by this scheme.

Quality based frame selection for face clustering in news video

Clustering identities in a broadcast video is a useful task to aid in video annotation and retrieval. Quality based frame selection is a crucial task in video face clustering, to both improve the clustering performance and reduce the computational cost. We present a frame work that selects the highest quality frames available in a video to cluster the face. This frame selection technique is based on low level and high level features (face symmetry, sharpness, contrast and brightness) to select the highest quality facial images available in a face sequence for clustering. We also consider the temporal distribution of the faces to ensure that selected faces are taken at times distributed throughout the sequence. Normalized feature scores are fused and frames with high quality scores are used in a Local Gabor Binary Pattern Histogram Sequence based face clustering system. We present a news video database to evaluate the clustering system performance. Experiments on the newly created news database show that the proposed method selects the best quality face images in the video sequence, resulting in improved clustering performance.

Structural layers of communication on Twitter

Twitter is used for a range of communicative purposes. These extend from personal tweets that address what used to be Twitter’s default question, “What’s happening?”, through one-on-one @reply conversations between close friends and attempts at getting the attention of celebrities and other public actors, to discussions in communities built around specific issues—and back again to broadcast-style statements from well-known individuals and brands to their potentially very large retinue of followers.

Predicting shot locations in tennis using spatiotemporal data

Over the past decade, vision-based tracking systems have been successfully deployed in professional sports such as tennis and cricket for enhanced broadcast visualizations as well as aiding umpiring decisions. Despite the high-level of accuracy of the tracking systems and the sheer volume of spatiotemporal data they generate, the use of this high quality data for quantitative player performance and prediction has been lacking. In this paper, we present a method which predicts the location of a future shot based on the spatiotemporal parameters of the incoming shots (i.e. shot speed, location, angle and feet location) from such a vision system. Having the ability to accurately predict future short-term events has enormous implications in the area of automatic sports broadcasting in addition to coaching and commentary domains. Using Hawk-Eye data from the 2012 Australian Open Men's draw, we utilize a Dynamic Bayesian Network to model player behaviors and use an online model adaptation method to match the player's behavior to enhance shot predictability. To show the utility of our approach, we analyze the shot predictability of the top 3 players seeds in the tournament (Djokovic, Federer and Nadal) as they played the most amounts of games.

Orchids : My Intersex Adventure

Orchids: My Intersex Adventure is a multi-award winning autobiographical documentary film. The film follows documentary filmmaker, Phoebe Hart, as she comes clean on her journey of self-discovery to embrace her future and reconcile the past shame and family secrecy surrounding her intersex condition. Despite her mother’s outright refusal to be in the film, Phoebe decides she must push on with her quest to resolve her life story and connect with other intersex people on camera. With the help of her sister Bonnie and support from her partner James, she hits the open road and reflects on her youth. Phoebe’s happy and carefree childhood came to an abrupt end at puberty when she was told she would never menstruate nor have children. But the reasons why were never discussed and the topic was taboo. At the age of 17, Phoebe’s mother felt she was old enough to understand the true nature of her body and the family secret was finally revealed. Phoebe then faced an orchidectomy, invasive surgery to remove her undescended testes, the emotional scars of which are still raw today. Phoebe’s road trip around Australia exposes her to the stories of other intersex people and holds a mirror to her own experience. She learns valuable lessons in resilience and healing but also sees the pervasive impact her condition has on all her relationships. At home, Phoebe and James want to start a family but dealing with infertility and the stress of the adoption process puts pressure on their marriage. Phoebe also starts to understand the difficult decisions her parents faced and is excited but apprehensive when they eventually agree to be interviewed. Will talking openly with her mother give Phoebe the answers she has been looking for? The film was produced and directed by Phoebe Hart and commissioned by the Australian Broadcasting Commission. The film premiered at the Brisbane International Film Festival in 2010 where it was voted the number one film of the festival by audiences. Orchids was broadcast on ABC1 in Australia in 2012, appeared in more than 50 film festivals internationally and has since been broadcast nationally in Switzerland, Sweden, Israel, Spain, France, Russia, Poland, Germany and the USA.

Broadcasting digital stories on Australian community television : the Digistories Anthology

Community-based arts and media movements have been intsrumental in building population-wide creative capacity for cultural development, social participation and social transformation in many parts of the world. Digital storytelling is a form of media practice that was pioneered in the United States at the intersection of these movements. It is described here as a ‘co-creative’ media production method. This description aims to differentiate the approaches to collaborative content creation that are used in community cultural development (CCD) and community media movements from those valued in professional and consumer modes of media production. Yet, the products of co-creative practices, such as digital stories, do not circulate widely through existing media networks or through the newer social media networks that Australian CCD and community media movements anticipated by at least twenty years. The complex politics of story ownership are one of a number of factors that often render ‘publication’ a secondary consideration in the making of digital stories. The possibility of ‘downstream’ use and re-use of stories in other networks is not usually considered in initial planning and development processes. As landmark projects such as Capture Wales indicate, even where stories are made for broadcast outcomes, television can be a problematic window for exhibiting digital stories. Scepticism about the brave new world of reality television and user generated content also circulates in digital storytelling networks, especially when it comes to ethical concerns for managing the risks of harm associated with widespread distribution of digital stories to indiscriminate publics. This publication reports on a collaborative action research project that took a closer look at some of the constraints relating to the problems of re-purposing digital stories for television. It focussed on ‘best practice’ for managing the risks of harm to storytellers in the process of re-purposing digital stories for broadcast on community television.

The art of co-creative media : an Australian survey

This article describes how - in the processes of responding to participatory storytelling practices - community, public service, and to a lesser extent, commercial media institutions are themselves negotiated and changed. Although there are significant variations in the conditions, durability, extent, motivations and quality of these developments and their impacts, they nonetheless increase the possibilities and pathways of participatory media culture. This description first frames digital storytelling as a ‘co-creative’ media practice. It then discusses the role of community arts and cultural development (CACD) practitioners and networks as co-creative media intermediaries, and then considers their influence in Australian broadcast and Internet media. It looks at how participatory storytelling methods are evolving in the Australian context and explores some of the implications for cultural inclusion arising from a shared interest in ‘co-creative’ media methods and approaches.

Making orchids – Gardening an intersex experience on videotape [In German - Orchideen Züchten. Eine inter Erfahrung auf Film]

In 2004, my thirtieth year of life, I began to develop and produce a documentary about the lived experience of being intersex. At the time, I didn’t ever expect the film would be autobiographical in nature. I’d known I was intersex since I was 17, and aware of my difference for many years prior, and I’d been making and presenting documentaries for almost as long, yet the idea to expose myself so publicly was frightening to me. However, I realised I couldn’t expect others to step in front of the lens when I didn’t have the courage to do so myself. The final result was Orchids: My Intersex Adventure, which maps my intersex journey from shame, stigma and secrecy to self‐acceptance. The film has now been broadcast on television sets around the world. It has also won many awards and appeared in numerous film festivals....

Too many cooks? Co-creating and co-teaching studio courses in the creative media context : a career-focussed approach

This Case Study relates to the creation and implementation of career‐focussed courses in Creative Media for film, television, animation, broadcast and web contexts. The paper examines the advantages and disadvantages of co‐teaching, and how different professional and academic backgrounds and disciplines can productively inform curriculum design and delivery in the academic/professional context. The authors, as co‐creators and co‐lecturers, have developed a number of courses which represent current working models for intermediate to advanced level academic/professional study, and attract students from across the creative disciplines; including theatre, media, visual arts and music. These courses are structured to develop in students a wide range of aesthetic and technical skills, as well as their ability to apply those skills professionally within and across the creative media industries. Issues regarding the balance between academic rigour, practical hands‐on skill development, assessment, logistics, resources, teamwork and other issues, are examined in the paper.