Corpus ID: 236171233

AnonySIGN: Novel Human Appearance Synthesis for Sign Language Video Anonymisation

  title={AnonySIGN: Novel Human Appearance Synthesis for Sign Language Video Anonymisation},
  author={Ben Saunders and Necati Cihan Camgoz and R. Bowden},
The visual anonymisation of sign language data is an essential task to address privacy concerns raised by largescale dataset collection. Previous anonymisation techniques have either significantly affected sign comprehension or required manual, labour-intensive work. In this paper, we formally introduce the task of Sign Language Video Anonymisation (SLVA) as an automatic method to anonymise the visual appearance of a sign language video whilst retaining the meaning of the original sign language… Expand

Figures and Tables from this paper

Mixed SIGNals: Sign Language Production via a Mixture of Motion Primitives
A novel Mixture of Motion Primitives (MOMP) architecture for sign language animation, where a set of distinct motion primitives are learnt during training, that can be temporally combined at inference to animate continuous sign language sequences. Expand


Everybody Sign Now: Translating Spoken Language to Photo Realistic Sign Language Video
This paper proposes SignGAN, the first SLP model to produce photo-realistic continuous sign language videos directly from spoken language, and introduces a method for controllable video generation, enabling training on large, diverse sign language datasets and providing the ability to control the signer appearance at inference. Expand
Text2Sign: Towards Sign Language Production Using Neural Machine Translation and Generative Adversarial Networks
This work presents a novel approach to automatic Sign Language Production using recent developments in Neural Machine Translation, Generative Adversarial Networks, and motion generation to produce sign videos from spoken language sentences. Expand
Adversarial Training for Multi-Channel Sign Language Production
An adversarial discriminator evaluates the realism of sign production conditioned on the source text, pushing the generator towards a realistic and articulate output, and set new benchmarks for the production of multi-channel sign to underpin future research into realistic SLP. Expand
Using a Language Technology Infrastructure for German in order to Anonymize German Sign Language Corpus Data
This work combines analysis of all data that the DGS Korpus (German Sign Language corpus) project has, including named entity recognition on translations into German, and uses the WebLicht language technology infrastructure. Expand
Neural Sign Language Synthesis: Words Are Our Glosses
  • J. Zelinka, J. Kanis
  • Computer Science
  • 2020 IEEE Winter Conference on Applications of Computer Vision (WACV)
  • 2020
The main goal in this paper was to design a fully end-to-end automatic sign language synthesis system trained only on available free data (daily TV broadcasting) and excluded any manual video annotation. Expand
Neural Sign Language Translation
This work formalizes SLT in the framework of Neural Machine Translation (NMT) for both end-to-end and pretrained settings (using expert knowledge) and allows to jointly learn the spatial representations, the underlying language model, and the mapping between sign and spoken language. Expand
Exploring Collection of Sign Language Datasets: Privacy, Participation, and Model Performance
The results suggest that privacy concerns may exist in contributing to sign language corpora, that filters (especially expressive avatars and blurred faces) may impact willingness to participate, and that training on more filtered data may boost recognition accuracy in some cases. Expand
Sign Language Transformers: Joint End-to-End Sign Language Recognition and Translation
A novel transformer based architecture that jointly learns Continuous Sign Language Recognition and Translation while being trainable in an end-to-end manner is introduced by using a Connectionist Temporal Classification (CTC) loss to bind the recognition and translation problems into a single unified architecture. Expand
Deep Gesture Video Generation With Learning on Regions of Interest
Experimental results show that the method can produce gesture videos with more realistic appearance and precise meaning than the state-of-the-art video generation approaches. Expand
Approaches to the Anonymisation of Sign Language Corpora
The motivations behind anonymisation and the close connection with the issue of ethics and informed consent for corpus participants are explored and the variety of ways in which these can be carried out are described. Expand