FIAT/IFTA World Conference 2024

Considering the growing popularity of podcasts due to their inherent convenience for consumption in areas such as deepening knowledge and entertainment, and taking the example of the Rai Play Sound platform, where original podcasts can be found, in addition to the radio offering of new and/or repertoire programs, we will analyse, through concrete examples, the methods and verification processes implemented by Rai’s Rights Management Department when rights clearances are requested in order to use archive material in new in-house or third parties podcast productions.

We will examine the different types of contracts, from the oldest ones, which only provided for radio and television rights, to the newest ones where exploitation within podcasts is already provided for in the contractual clauses, explaining what risks we take and why. Rights Management Department follows Rai’s Legal Department’s guidelines , but its decisions are also the result of comparisons and evaluations made for each individual case based on various factors: whether internal use or commercialization of the podcast is envisaged; the nature of the repertoire requested (movies, entertainment, investigations, sports, etc.); whether the podcast consists of the full retransmission of a program already broadcasted or whether the material must be inserted into a new work.

We will also discuss cases where the Fair Use principle can be applied for the free use of archives (for example, for purposes of criticism, comment, information, teaching, instruction, or research).

During the presentation, we will show some specific and potentially critical cases, discussing how we behave in the presence of material protected by copyright and agreements between Rai and collecting societies.

This session will explore how the BFI approached its largest ever rights clearance project, researching over 30,000 titles and clearing rights for 13,000 film and TV works for a new streaming service in UK libraries, BFI Replay, as part of the recently completed Heritage 2022 programme. I'll show how we devised our approach to compliance & risk across the multiple layers of rights and the different categories of rights holders and works, and the tools, processes and policies we developed to support this work and the results we achieved. I'll share some of the main obstacles we faced, like the information gap, and how we engaged with rights holders and consider some of the longer term challenges that can be created by project driven rights work.

INA is a French public company whose first mission is to preserve and develop a vast collection of audiovisual and audio archives from French public radios and TV channels.

Due to the quantity of rights managed and the large number of rightsholders in the collection of INA’s archives, exploiting it involves taking risks.

In order to limit risk exposure, INA has developed a politic of collective agreement conclusion with CMOs (for the authors rights) and unions (for the performers rights). In the same time, individual agreements are concluded / negotiated with third-party property rights (rightsholders non-members of CMOs, coproducers, organizers of sports events / shows, multiple rightsholders of contents previously incorporated in INA’s archives (photo, phonograms, literary works, etc.).

INA’s Legal Department also have to deal with rights of the personality, image, freedom of expression, private life… and others (with the IA, RGPD).

Our purpose is to facilitate the work of our operational teams with clear guidelines on each subject. A legal team at the INA’s Legal Department is dedicated to the legal analysis of our contents with the perspective of an exploitation to manage the risk. INA has launched a project to create a SI which enables to combine the legal data on the contents and the royalties data in order to simplify the decision making. However, human intervention is still necessary to appreciate the risk and the opportunity of an exploitation.

In 2022, Studio Silverback, Britain's leading natural history producers, proposed creating an online library of world-class footage of our changing planet for use in non-commercial climate and environmental projects and communications. The goal was to build a high-quality, scientifically accurate, and globally accessible collection of video assets, to be named "Open Planet", and make it available for free. This idea stemmed from frustration with broadcasters and streamers tightly controlling the rights to Silverback's documentary content, including unused rushes material, restricting its use by others.

Following a rapid project timeline which addressed investment, corporate structures, technological, rights and media management requirements, the beta library was launched in Autumn 2023 and has already made a significant impact. Footage has been used by a mix of audiences – from scientists and educators, to NGOs, activists and filmmakers – in multi-channel communications campaigns and events, including major global conferences such as COP28, the UN General Assembly and the World Economic Forum’s Annual Meeting.

The full global launch of Open Planet is scheduled for Spring 2024. In this presentation, Rights and Archive consultants from Northbound will explain the construction of the library and its current reach. Topics will include:

• Technology procurement and system build 
• Media management and metadata model 
• Rights negotiations and unlocking 
• Usage and impacts of the library 

We will present results of combining the latest research in automatic speech recognition (ASR) with European high performance computing (HPC) and large quantities of raw audiovisual data contained in national radio and television archives. The aim of the work was two-fold, firstly, to advance ASR by building models on large public data collections and secondly, to harness the large audio-visual media archives for large-scale qualitative and quantitative media research by generating an automatic indexation based on all spoken content that is decoded by ASR. Only the largest global companies can have access both to the latest ASR development, huge computing resources and huge audio collections, but their commercial interests do not treat all languages equally.

In Europe, most languages are spoken in small countries which, however, have advanced radio and television archives containing millions of hours of broadcasted media content. The latest publicly funded HPC initiatives have also opened researchers an access to unprecedented computational resources. By utilizing the computing and archives it is possible for researchers to develop and publish large pre-trained speech models for many languages without depending on the commercial interests of the large global companies. The large speech models can be pre-trained in a self-supervised fashion which can benefit also from untranscribed and uncategorized audio collections. When openly published, these models make it then easy and quick to develop speech technology applications, such as accurate recognizers for ASR and speech, speaker and audio characteristics for these languages by fine-tuning the models using a feasible amount of transcribed target data.

In the past years, Large Language Models (LLMs) have been increasingly developed and used for various Natural Language Processing (NLP) tasks. As an example, LLMs can be employed for content classification without a need for manual human annotation or exhaustive training datasets. However, such LLMs are associated with high computational cost during inference, preventing their wide adoption in the audiovisual industry, such as large scale audiovisual collections.

In this context, we want to automatically classify broadcast news into topics. Topic classification can be useful for audiovisual content retrieval or monitoring, but remains a challenging problem, particularly due to the difficult task of segmenting continuous feed into homogenous extracts. To solve this issue, we propose a framework applied on French TV and radio news where we transcribe a dataset of 11.7k hours, broadcasted in 2023 on 21 channels with a State-of-the-Art transcription model. A LLM is used in few-shot conversation mode to obtain a topic classification on those transcriptions. We define a topic classification scheme based on the IPTC categories.

Using the generated LLM annotations, we explore the finetuning of a specialized smaller classification model. To evaluate the performances of these models, and to estimate the subjectivity of the topic categorization task, we construct and annotate a test set of 03h44m. We demonstrate that while LLM's inference costs makes them prohibitive for large scale analysis of audiovisual collections, they can be used to generate synthetic datasets used to train less complex models (students) associated to much smaller inference time and better classification performances.

Finally, using an automatic gender classification tool, we compute the speaking time per gender depending on the topic, to determine if some subjects are predominantly reserved to specific genders. We show that women are notably under-represented in subjects such as sports and politics.

The transformative potential of applying artificial intelligence in the context of audiovisual archiving has been demonstrated across a spectrum of use cases related to search & exploration, preservation, artistic expression and big-data analysis. This presentation explores research findings regarding the integration of AI solutions specifically in humanities and social sciences research.  

The emergence of Large Language Models (LLMs) and Large Multimedia Models (LMMs) poses new opportunities and questions in this area. For instance, LLMs could support the content annotation tasks for researchers working on framing analysis over large amounts of textual data. But more investigation is required to determine how such systems could be fine tuned and integrated into research workflows to deliver satisfactory and trustworthy outcomes. The goal of this research was to develop a roadmap that guides the development and use of LLMs and LMMs in SSH research. It (i) Identifies potential SSH research areas (such as framing analysis) that could benefit from the use of LMMs / LLMs, (ii) Identify limitations and risks of these techniques, (iii) provide recommendations on how LLMs and LMMs could be designed and fine tuned to fit SSH scholar needs.

We will also reflect on how researchers, archives, and the broader cultural sector can contribute to promoting more responsible AI practices in society and raising "AI Literacy." This entails exploring practical ways in which cultural institutions can actively engage with the ethical and societal implications of AI, such as curating exhibitions that examine AI's impact on human lives, hosting workshops to educate the public about AI technologies and their implications, and collaborating with researchers to develop resources that foster critical thinking about AI in diverse communities.

Daily archives or Archives of the day is a very popular format for accessing specific archive content. What happens when we start to use the “old” daily news as a trigger for intergenerational dialogue? Can we understand each other better through archival footage?

Three generations in particular, Baby Boomers, Generations X and Millennials, are currently meeting each other on Facebook. Each has a different (or none) experience of life under socialism in Czechoslovakia. Across the generations, there are also different views on whether things were better before and where the current society is heading. News events from the 1970s and 1980s evoke a wide range of emotions. Comments on the Czech Television Archive's Facebook page commonly include statements such as "the world was still fine then", but also those that condemn everything from the old days. But does it have to be so black and white?

We think primarily about the relationship between the archival material, the archivist who selects the content, and the audience. Where, how and under what conditions can everyone safely meet and develop these relationships? Can we use a selection of events from the archive, well-aimed questions and sensitively guided discussion to create a space for cultivated dialogue among our viewers? And what can the audience teach us? Why do they watch archival footage? What does it evoke in them? Where, apart from social networks, can we meet and what form to choose? If we open ourselves up to dialogue and resist nothing, we may discover what else archives have to offer society.

At the end of 2023, adapted from the Mao Dun Literature Award work "Blossoms", directed by Kar Wai Wong (HK), the Shanghai local television drama "Blossoms Shanghai" had been launched, with a great reputation. Shanghainese seem to have opened a memory gate, collectively returning to the vibrant 1990s. Many real-life scenes from the past appear in the drama, most of these historical images were provided by Shanghai Audio-Visual Archives.

In 2020, the production team of "Blossoms Shanghai" had contact with Shanghai Audio-Visual Archives. With the aid of the archives of the city of Shanghai in the 1980s to 1990s, they hope to dig deeper into the special historical period of Shanghai and the living conditions of its people. Also, they hope to realistically recreate the scene of the drama at that time.

In the following three years, the Archives provided senior professional editors and researchers with specialized consulting services for the production team, with dedicated personnel conducting targeted queries and organizing information. They provided over 60 historical video archives for the production team of "Blossoms Shanghai", totalling over 1800 minutes. The production team of "Blossoms Shanghai" utilized these rich visual memories to create a 1:1 replica of the street scenes of the old Shanghai at the film and television base. The narrow alleyways, crowded staircases, and aisles filled with things all present the rich fireworks and contemporary atmosphere of Shanghai in the early 1990s in the drama.

From the initial historical scene reference needs of the cooperation to the copyright purchase of precious historical images for drama applications, the Shanghai Audio-Visual Archives and the production team have worked together to meticulously craft a sword for three years, showcasing the touching urban Shanghai of the 1990s. This form of market-oriented application has been continuously carried out by Shanghai Audio-Visual Archives for several years, providing services for multiple dramas, such as the scientist series film "Ye Shuhua" and "Yan Dongsheng", the drama "Great Pudong", " Like a Flowing River", "Faith Makes Great", and so on.

Real historical images not only serve as an important reference for the production team, but also make the content presented in movies and TV dramas more convincing by examining details. In addition, the direct use of real historical images in the drama reflects the historical background while making the content of the film and television drama more creative. In the face of more diversified media communication methods, relying on the collection of historical images, Shanghai Audio-Visual Archives has also planned and produced a variety of short videos on themes, showing the audience the real historical images at that time on new media platforms such as the official WeChat channel and the official account, which has made many people relive the past life, popularize historical knowledge, and gradually build a social public library.

The state television series, ‘Yuruparí: Popular Traditional Art’, documented cultural expressions of peasant, Afro and indigenous populations in Colombia, leaving a memory of the sociopolitical configurations in the midst of the armed conflict of the 1980s. Being the most complete ethnographic record of its time, with rituals, festivals and songs declared Intangible Heritage of Humanity by UNESCO, the preservation of this collection after its censorship, invites us to reflect on access and restitution policies, in order to unveil untold stories and embedded narratives.

Following the beginning of its restoration in 2013 led by the colombian institutions of Proimágenes Colombia, RTVC-Señal Memoria and Fundación Patrimonio Fílmico Colombiano, as well as the lead researcher Juan Suárez PhD and international organizations such as the National Museum of African American Culture-Smithsonian/Blake McDowell, FIAT/IFTA, amongst others, 24 of the 86 episodes in 16mm have been completed.

The central figure of Gloria Triana, its main director, has led the conventional narrative around this archive. However, based on the documentation management and contextualization with a gender perspective of the private, state and community archives around this collection, more than 25 participating women who constitute part of Colombian audiovisual history have been made visible; Ann Marie Lóök, Beatriz Barros and María Ema Frade are just some of them.

In this way, in an exercise to recover the oral memory of the production team, the participation of the populations represented in the collective cataloging of their episodes, and the making of a film with the found materials, this project proposes different approaches to reread the sociopolitical narrative(s) imposed in the series and delve into the challenges of access, reuse and promotion of public archives in Colombia.

Despite considerable research on explicitly sexualized talk as a form of gender micro-aggression, hostility and harassment, it is still unclear how does talking about less-explicitly sexualizing but no less gendered issue such as beauty, shape the paralinguistics of interaction. Taking the topic of human beauty as our case study and utilizing short moments from TV and film archives, we therefore ask:

1) How does talking about an explicitly gendered issue (beauty) contribute to the overall gendering of audiovisual cultural objects?

2) How do the paralinguistic aspects of audiovisual gendered interactions shift overtime?

This proof of concept (POC) research is testing the usefulness of experimental AI methods to reveal the multilayered effect of gender on interactions. The POC consists of two one-on-one interviews in which beauty is a specifically mentioned topic, with two women beauty queens. The two samples represent two periods of “beauty talk” – one from the 1970ies and one from the 2010s.

Following the design of a set of coding labels, we will index two videos manually and then compare the qualitative process to an automatic indexing of these videos using an AI indexing tool. Among its analytical insights, we will automatically extract multimodal information and insights such as automatic transcription of spoken Hebrew; Speaker identification, vocal effects (such as laughter or long silences), sentiment analysis, and more.

The comparison will help identify specific spoken/visual/vocal profiles of both interviewers and interviewees, as well as of the interactions between them and to set the Beauty features that would be included in the training of a future study, on a large scale database of interviews. Using both human and AI labeling, this POC research hopes to contribute to a more precise hermeneutic interpretations of the human beauty concept and to enable a “distant reading, viewing, and listening” of these multimodal artifacts.

During the footage research for the production of a documentary about sexism in the world of sport, it is highlighted how not only the audiovisual documents themselves but also the metadata that accompany them reflect the model of society and the framework of thought at the moment in which they were created. The challenge of the research was to find fragments that showed attitudes that attract attention today for their sexist tone but that at the time were considered normal and therefore were not described as such in the metadata. The research had to be done blindly, listening to interviews, anticipating the results and thinking about how they would have been described years ago, appealing to the memory of the oldest workers in the medium and navigating the ancient thesaurus, which is in itself a reflection of this thinking 40 years ago when women's sport was not at the same level as men's.  

Once the material has been located and the documentary made, the following question arises: Should we or should we not correct the metadata in our MAM system and update the description according to the current parameters? Our archive has the dual mission of preserving the audiovisual heritage while serving current broadcasting. To fulfill the first mission shouldn't we keep the material with the original metadata? Whereas, to fulfill the second mission, shouldn't we update them to facilitate research for the production of new material?

We will accompany the presentation of a small sample of the material found during the research to exemplify the need to revise the metadata to make these items accessible for future productions.

The paper titled "Archive Iraq: Films are rising again. How, when and who?" by Wareth Kwaish, a Project Officer of the Iraqi Archive Project, discusses the efforts to preserve Iraq's cinematic heritage. It highlights the historical richness of Iraqi visual memory, emphasizing the importance of films as cultural, historical, and social treasures.

The paper outlines the establishment of the Committee for the Iraqi Visual Memory, supported by the French Embassy and Expertise France, to restore and digitize over 100 Iraqi films. The project involves training Iraqi professionals in film assessment, inventory, restoration, and digitization techniques.

Objectives include creating a digital film library accessible for education and research, raising societal awareness on film preservation, and preserving Iraq's cultural legacy. The methodology encompasses film selection, evaluation, restoration, and digitization, coupled with community engagement. The project aims to ensure the continuity of Iraqi visual memory, contributing to the larger Iraq Archive Project, preserving the nation's heritage for future generations and global appreciation.

In the contemporary landscape of news media, we encounter a realm where simulation and stereotypes pervade, shaping our perception of reality. Understanding the visual language of news media is crucial for unraveling how representations are meticulously crafted. It's posited that the visual narrative across various media platforms has become increasingly uniform, a trend significantly influenced by the reliance on stock shots. This practice has seen a noticeable uptick in the last two decades, paralleling the archival turn in media production. Rather than consistently sourcing new material, the efficiency, cost-effectiveness, and digital advancements have led journalists to frequently repurpose stock shots for visual narratives in their reporting.

The ANR CROBORA project, in collaboration with prominent entities like Ina, Rai Radiotelevisione Italiana, and Mediaset, seeks to cast light on the repetitive and emblematic use of stock shots in audiovisual media. By amassing a collection of 35,000 stock footage reflecting European integration from two decades of French and Italian television broadcasts and selected web narratives, the project endeavors to analyze this corpus of mediatized visual memory. A sophisticated visualization platform and analytical tools have been developed to navigate this extensive and diverse dataset, addressing the challenges of collecting, organizing, annotating, and visualizing a large-scale and varied collection of audiovisual content.

This communication aims to delve deeper into the methodologies of data collection, annotation, and explorative analysis, leveraging both metadata and the audiovisual data itself. Through our data-driven approach, our research highlights the construction of collective imagination through the media's repetitive imagery, investigating how audiovisual production reinforces stereotypes, thereby influencing cultural dynamics and visual culture.

The mass-scale digitisation project Images for the Future project, concluded in 2014, allowed an important part of Dutch audiovisual heritage to be digitised and preserved. The promised land of wide accessibility for the public however could not be reached at that time due to copyright restrictions limiting the re-exposure of our archival treasures.  

Since the EU Copyright Directive introduced mechanisms that allow for out-of-commerce works to be published on cultural heritage portals, unless rights holders opt-out, Sound & Vision started to prepare the potential mass-scale publication of content. However, incomplete, erroneous, or conflicting information complicates the selection of material suitable for public release. Additionally, inconsistent metadata quality and lack of standardization for diverse audiences obstruct efficient and appropriate online publication. In this presentation the audience is invited at Sound & Visions kitchen table to take a closer look at how we analyse collection-assets’ data to establish (i) if they can be labled as an Out of Commerce work, (ii) if they are technically ready for publication and (iii) what metadata changes are potentially needed in order to present them in an understandable way to a wider audience.

We propose a brief presentation of four conceptual elements that may underpin an automated cataloging process of archival materials, summarizing them through a free adaptation of the biblical Commandments. The chosen ones are:
 
• Thou shalt describe events about all things.
• Thou shalt univocal identify the event in linked properties and relationships.
• Thou shalt use equivalences to transform the captured external data into your metadata  standards.
• Thou shalt describe the documents in cascade.
 
These four theoretical pillars arise from an idea that challenges the traditional approach to the documentary process by reversing the object of analysis, which almost always has focused on the document. Therefore, the first and main commandment relies on this thought, and encourages us  to do it the other way around,  going from the analysis of the documents that report events to analyse events captured in documents... And let's see what happens.
 
The second commandment, and based on the experience of the semantic web, indicates that events must be defined as unique concepts described in properties and relationships, which are linked to other concepts contained in controlled vocabularies.
 
The third statement focuses on the capture of external data applying the principle of equivalence based on the transformation of these data into metadata with the desired standardized form in order to assign each concept its equivalent in the field where it is represented.
 
And finally, the fourth commandment suggests the process of describing documents in cascade mode. That is, to use the common metadata of an event and replicate them in all the documents that refer to it.