Sign language representation learning presents unique challenges due to the complex spatio-temporal nature of signs and the scarcity of labeled datasets. Existing methods often rely either on models pre-trained on general visual tasks, that lack sign-specific features, or use complex multimodal and multi-branch architectures. To bridge this gap, we introduce a scalable, self-supervised framework for sign representation learning. We leverage important inductive (sign) priors during the training of our RGB model. To do this, we leverage simple but important cues based on skeletons while pretraining a masked autoencoder. These sign specific priors alongside feature regularization and an adversarial style agnostic loss provide a powerful backbone. Notably, our model does not require skeletal keypoints during inference, avoiding the limitations of keypoint-based models during downstream tasks. When finetuned, we achieve state-of-the-art performance for sign recognition on the WLASL, ASL-Citizen and NMFs-CSL datasets, using a simpler architecture and with only a single-modality. Beyond recognition, our frozen model excels in sign dictionary retrieval and sign translation, surpassing standard MAE pretraining and skeletal-based representations in retrieval. It also reduces computational costs for training existing sign translation models while maintaining strong performance on Phoenix2014T, CSL-Daily and How2Sign.

A major impediment to the advancement of sign language translation (SLT) is data scarcity.Much of the sign language data currently available on the web cannot be used for training supervised models due to the lack of aligned captions.Furthermore, scaling SLT using large-scale web-scraped datasets bears privacy risks due to the presence of biometric information, which the responsible development of SLT technologies should account for.In this work, we propose a twostage framework for privacy-aware SLT at scale that addresses both of these issues.We introduce SSVP-SLT, which leverages selfsupervised video pretraining on anonymized and unannotated videos, followed by supervised SLT finetuning on a curated parallel dataset.SSVP-SLT achieves state-of-the-art finetuned and zero-shot gloss-free SLT performance on the How2Sign dataset, outperforming the strongest respective baselines by over 3 BLEU-4.Based on controlled experiments, we further discuss the advantages and limitations of self-supervised pretraining and anonymization via facial obfuscation for SLT. 1

AR/VR devices have started to adopt hand tracking, in lieu of controllers, to support user interaction. However, today’s hand input rely primarily on one gesture: pinch. Moreover, current mappings of hand motion to use cases like VR locomotion and content scrolling involve more complex and larger arm motions than joystick or trackpad usage. STMG increases the gesture space by recognizing additional small thumb-based microgestures from skeletal tracking running on a headset. We take a machine learning approach and achieve a 95.1% recognition accuracy across seven thumb gestures performed on the index finger surface: four directional thumb swipes (left, right, forward, backward), thumb tap, and fingertip pinch start and pinch end. We detail the components to our machine learning pipeline and highlight our design decisions and lessons learned in producing a well generalized model. We then demonstrate how these microgestures simplify and reduce arm motions for hand-based locomotion and scrolling interactions.

Automatic Sign Language Translation requires the integration of both computer vision and natural language processing to effectively bridge the communication gap between sign and spoken languages. However, the deficiency in large-scale training data to support sign language translation means we need to leverage resources from spoken language. We introduce, Sign2GPT, a novel framework for sign language translation that utilizes large-scale pretrained vision and language models via lightweight adapters for gloss-free sign language translation. The lightweight adapters are crucial for sign language translation, due to the constraints imposed by limited dataset sizes and the computational requirements when training with long sign videos. We also propose a novel pretraining strategy that directs our encoder to learn sign representations from automatically extracted pseudo-glosses without requiring gloss order information or annotations. We evaluate our approach on two public benchmark sign language translation datasets, namely RWTH-PHOENIX-Weather 2014T and CSL-Daily, and improve on state-of-the-art gloss-free translation performance with a significant margin.

This paper is a brief summary of the First WMT Shared Task on Sign Language Translation (WMT-SLT22), a project partly funded by EAMT. The focus of this shared task is automatic translation between signed and spoken languages. Details can be found on our website (https://www.wmt-slt.com/) or in the findings paper (Müller et al., 2022).

This paper investigates data-free class-incremental learning (DFCIL) for hand gesture recognition from 3D skeleton sequences. In this class-incremental learning (CIL) setting, while incrementally registering the new classes, we do not have access to the training samples (i.e. data-free) of the already known classes due to privacy. Existing DFCIL methods primarily focus on various forms of knowledge distillation for model inversion to mitigate catastrophic forgetting. Unlike SOTA methods, we delve deeper into the choice of the best samples for inversion. Inspired by the well-grounded theory of max-margin classification, we find that the best samples tend to lie close to the approximate decision boundary within a reasonable margin. To this end, we propose BOAT-MI – a simple and effective boundary-aware prototypical sampling mechanism for model inversion for DFCIL. Our sampling scheme outperforms SOTA methods significantly on two 3D skeleton gesture datasets, the publicly available SHREC 2017, and EgoGesture3D – which we extract from a publicly available RGBD dataset. Both our codebase and the EgoGesture3D skeleton dataset are publicly available: https://github.com/humansensinglab/dfcil-hgr.

Sign languages are visual languages, with vocabularies as rich as their spoken language counterparts. However, current deep-learning based Sign Language Production (SLP) models produce under-articulated skeleton pose sequences from constrained vocabularies and this limits applicability. To be understandable and accepted by the deaf, an automatic SLP system must be able to generate co-articulated photo-realistic signing sequences for large domains of discourse. In this work, we tackle large-scale SLP by learning to co-articulate between dictionary signs, a method capable of producing smooth signing while scaling to unconstrained domains of discourse. To learn sign co-articulation, we propose a novel Frame Selection Network (FS-NET) that improves the temporal alignment of interpolated dictionary signs to continuous signing sequences. Additionally, we propose SIGNGAN, a pose-conditioned human synthesis model that produces photo-realistic sign language videos direct from skeleton pose. We propose a novel keypoint-based loss function which improves the quality of synthe-sized hand images. We evaluate our SLP model on the large-scale meineDGS (mDGS) corpus, conducting extensive user evaluation showing our FS-NET approach improves coarticulation of interpolated dictionary signs. Additionally, we show that SIGNGAN significantly outperforms all baseline methods for quantitative metrics, human perceptual studies and native deaf signer comprehension.

Disentangled representations support a range of downstream tasks including causal reasoning, generative modeling, and fair machine learning. Unfortunately, disentanglement has been shown to be impossible without the incorporation of supervision or inductive bias. Given that supervision is often expensive or infeasible to acquire, we choose to incorporate structural inductive bias and present an unsupervised, deep State-Space-Model for Video Disentanglement (VDSM). The model disentangles latent time-varying and dynamic factors via the incorporation of hierarchical structure with a dynamic prior and a Mixture of Experts decoder. VDSM learns separate disentangled representations for the identity of the object or person in the video, and for the action being performed. We evaluate VDSM across a range of qualitative and quantitative tasks including identity and dynamics transfer, sequence generation, Fréchet Inception Distance, and factor classification. VDSM achieves state-of-the-art performance and exceeds adversarial methods, even when the methods use additional supervision.

Computational sign language research lacks the large-scale datasets that enables the creation of useful real-life applications. To date, most research has been limited to prototype systems on small domains of discourse, e.g. weather forecasts. To address this issue and to push the field forward, we release six datasets comprised of 190 hours of footage on the larger domain of news. From this, 20 hours of footage have been annotated by Deaf experts and interpreters and is made publicly available for research purposes. In this paper, we share the dataset collection process and tools developed to enable the alignment of sign language video and subtitles, as well as baseline translation results to underpin future research.

In this paper, we focus on the task of one-shot sign spotting, i.e. given an example of an isolated sign (query), we want to identify whether/where this sign appears in a continuous, co-articulated sign language video (target). To achieve this goal, we propose a transformer-based network, called Sign-Lookup. We employ 3D Convolutional Neural Networks (CNNs) to extract spatio-temporal representations from video clips. To solve the temporal scale discrepancies between the query and the target videos, we construct multiple queries from a single video clip using different frame-level strides. Self-attention is applied across these query clips to simulate a continuous scale space. We also utilize another self-attention module on the target video to learn the contextual within the sequence. Finally mutual-attention is used to match the temporal scales to localize the query within the target sequence. Extensive experiments demonstrate that the proposed approach can not only reliably identify isolated signs in continuous videos, regardless of the signers' appearance, but can also generalize to different sign languages. By taking advantage of the attention mechanism and the adaptive features, our model achieves state-of-the-art performance on the sign spotting task with accuracy as high as 96% on challenging benchmark datasets and significantly outperforming other approaches.

The visual anonymisation of sign language data is an essential task to address privacy concerns raised by large-scale 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 sequence. To tackle SLVA, we propose Anonysign, a novel automatic approach for visual anonymisation of sign language data. We first extract pose information from the source video to remove the original signer appearance. We next generate a photo-realistic sign language video of a novel appearance from the pose sequence, using image-to-image translation methods in a conditional variational autoencoder framework. An approximate posterior style distribution is learnt, which can be sampled from to synthesise novel human appearances. In addition, we propose a novel style loss that ensures style consistency in the anonymised sign language videos. We evaluate ANONYSIGN for the SLVA task with extensive quantitative and qualitative experiments highlighting both realism and anonymity of our novel human appearance synthesis. In addition, we formalise an anonymity perceptual study as an evaluation criteria for the SLVA task and showcase that video anonymisation using ANONYSIGN retains the original sign language content.

In this work, we propose a framework to collect a large-scale, diverse sign language dataset that can be used to train automatic sign language recognition models.The first contribution of this work is SDTrack, a generic method for signer tracking and diarisation in the wild. Our second contribution is SeeHear, a dataset of 90 hours of British Sign Language (BSL) content featuring more than 1000 signers, and including interviews, monologues and debates. Using SDTrack, the SeeHear dataset is annotated with 35K active signing tracks, with corresponding signer identities and subtitles, and 40K automatically localised sign labels. As a third contribution, we provide benchmarks for signer diarisation and sign recognition on SeeHear.

It is common practice to represent spoken languages at their phonetic level. However, for sign languages, this implies breaking motion into its constituent motion primitives. Avatar based Sign Language Production (SLP) has traditionally done just this, building up animation from sequences of hand motions, shapes and facial expressions. However, more recent deep learning based solutions to SLP have tackled the problem using a single network that estimates the full skeletal structure.We propose splitting the SLP task into two distinct jointly-trained sub-tasks. The first translation sub-task translates from spoken language to a latent sign language representation, with gloss supervision. Subsequently, the animation sub-task aims to produce expressive sign language sequences that closely resemble the learnt spatio-temporal representation. Using a progressive transformer for the translation sub-task, we propose a novel Mixture of Motion Primitives (MOMP) architecture for sign language animation. A set of distinct motion primitives are learnt during training, that can be temporally combined at inference to animate continuous sign language sequences.We evaluate on the challenging RWTH-PHOENIX-Weather-2014T (PHOENIX14T) dataset, presenting extensive ablation studies and showing that MOMP outperforms baselines in user evaluations. We achieve state-of-the-art back translation performance with an 11% improvement over competing results. Importantly, and for the first time, we showcase stronger performance for a full translation pipeline going from spoken language to sign, than from gloss to sign.

Abstract Sign languages are multi-channel visual languages, where signers use a continuous 3D space to communicate. Sign language production (SLP), the automatic translation from spoken to sign languages, must embody both the continuous articulation and full morphology of sign to be truly understandable by the Deaf community. Previous deep learning-based SLP works have produced only a concatenation of isolated signs focusing primarily on the manual features, leading to a robotic and non-expressive production. In this work, we propose a novel Progressive Transformer architecture, the first SLP model to translate from spoken language sentences to continuous 3D multi-channel sign pose sequences in an end-to-end manner. Our transformer network architecture introduces a counter decoding that enables variable length continuous sequence generation by tracking the production progress over time and predicting the end of sequence. We present extensive data augmentation techniques to reduce prediction drift, alongside an adversarial training regime and a mixture density network (MDN) formulation to produce realistic and expressive sign pose sequences. We propose a back translation evaluation mechanism for SLP, presenting benchmark quantitative results on the challenging PHOENIX14T dataset and setting baselines for future research. We further provide a user evaluation of our SLP model, to understand the Deaf reception of our sign pose productions.

Prior work on Sign Language Translation has shown that having a mid-level sign gloss representation (effectively recognizing the individual signs) improves the translation performance drastically. In fact, the current state-of-the-art in translation requires gloss level tokenization in order to work. We introduce a novel transformer based architecture that jointly learns Continuous Sign Language Recognition and Translation while being trainable in an end-to-end manner. This is achieved by using a Connectionist Temporal Classification (CTC) loss to bind the recognition and translation problems into a single unified architecture. This joint approach does not require any ground-truth timing information, simultaneously solving two co-dependant sequence-to-sequence learning problems and leads to significant performance gains. We evaluate the recognition and translation performances of our approaches on the challenging RWTH-PHOENIX-Weather-2014T (PHOENIX14T) dataset. We report state-of-the-art sign language recognition and translation results achieved by our Sign Language Transformers. Our translation networks outperform both sign video to spoken language and gloss to spoken language translation models, in some cases more than doubling the performance (9.58 vs. 21.80 BLEU-4 Score). We also share new baseline translation results using transformer networks for several other text-to-text sign language translation tasks.

Fair and unbiased machine learning is an important and active field of research, as decision processes are increasingly driven by models that learn from data. Unfortunately, any biases present in the data may be learned by the model, thereby inappropriately transferring that bias into the decision making process. We identify the connection between the task of bias reduction and that of isolating factors common between domains whilst encouraging domain specific invariance. To isolate the common factors we combine the theory of deep latent variable models with information bottleneck theory for scenarios whereby data may be naturally paired across domains and no additional supervision is required. The result is the Nested Variational AutoEncoder (NestedVAE). Two outer VAEs with shared weights attempt to reconstruct the input and infer a latent space, whilst a nested VAE attempts to reconstruct the latent representation of one image, from the latent representation of its paired image. In so doing, the nested VAE isolates the common latent factors/causes and becomes invariant to unwanted factors that are not shared between paired images. We also propose a new metric to provide a balanced method of evaluating consistency and classifier performance across domains which we refer to as the Adjusted Parity metric. An evaluation of NestedVAE on both domain and attribute invariance, change detection, and learning common factors for the prediction of biological sex demonstrates that NestedVAE significantly outperforms alternative methods.

Variational AutoEncoders (VAEs) provide a means to generate representational latent embeddings. Previous research has highlighted the benefits of achieving representations that are disentangled, particularly for downstream tasks. However, there is some debate about how to encourage disentanglement with VAEs, and evidence indicates that existing implementations do not achieve disentanglement consistently. The evaluation of how well a VAE's latent space has been disentangled is often evaluated against our subjective expectations of which attributes should be disentangled for a given problem. Therefore, by definition, we already have domain knowledge of what should be achieved and yet we use unsupervised approaches to achieve it. We propose a weakly-supervised approach that incorporates any available domain knowledge into the training process to form a Gated-VAE. The process involves partitioning the representational embedding and gating backpropagation. All partitions are utilised on the forward pass but gradients are backpropagated through different partitions according to selected image/target pairings. The approach can be used to modify existing VAE models such as beta-VAE, InfoVAE and DIP-VAE-II. Experiments demonstrate that using gated backpropagation, latent factors are represented in their intended partition. The approach is applied to images of faces for the purpose of disentangling head-pose from facial expression. Quantitative metrics show that using Gated-VAE improves average disentanglement, completeness and informativeness, as compared with un-gated implementations. Qualitative assessment of latent traversals demonstrate its disentanglement of head-pose from expression, even when only weak/noisy supervision is available.

Abstract We present a novel approach to automatic Sign Language Production using recent developments in Neural Machine Translation (NMT), Generative Adversarial Networks, and motion generation. Our system is capable of producing sign videos from spoken language sentences. Contrary to current approaches that are dependent on heavily annotated data, our approach requires minimal gloss and skeletal level annotations for training. We achieve this by breaking down the task into dedicated sub-processes. We first translate spoken language sentences into sign pose sequences by combining an NMT network with a Motion Graph. The resulting pose information is then used to condition a generative model that produces photo realistic sign language video sequences. This is the first approach to continuous sign video generation that does not use a classical graphical avatar. We evaluate the translation abilities of our approach on the PHOENIX14 T Sign Language Translation dataset. We set a baseline for text-to-gloss translation, reporting a BLEU-4 score of 16.34/15.26 on dev/test sets. We further demonstrate the video generation capabilities of our approach for both multi-signer and high-definition settings qualitatively and quantitatively using broadcast quality assessment metrics.

In this work we present a new approach to the field of weakly supervised learning in the video domain. Our method is relevant to sequence learning problems which can be split up into sub-problems that occur in parallel. Here, we experiment with sign language data. The approach exploits sequence constraints within each independent stream and combines them by explicitly imposing synchronisation points to make use of parallelism that all sub-problems share. We do this with multi-stream HMMs while adding intermediate synchronisation constraints among the streams. We embed powerful CNN-LSTM models in each HMM stream following the hybrid approach. This allows the discovery of attributes which on their own lack sufficient discriminative power to be identified. We apply the approach to the domain of sign language recognition exploiting the sequential parallelism to learn sign language, mouth shape and hand shape classifiers. We evaluate the classifiers on three publicly available benchmark data sets featuring challenging real-life sign language with over 1,000 classes, full sentence based lip-reading and articulated hand shape recognition on a fine-grained hand shape taxonomy featuring over 60 different hand shapes. We clearly outperform the state-of-the-art on all data sets and observe significantly faster convergence using the parallel alignment approach.

Sign language conveys information through multiple channels, such as hand shape, hand movement, and mouthing. Modeling this multichannel information is a highly challenging problem. In this paper, we elucidate the link between spoken language and sign language in terms of production phenomenon and perception phenomenon. Through this link we show that hidden Markov model-based approaches developed to model "articulatory" features for spoken language processing can be exploited to model the multichannel information inherent in sign language for sign language processing.

A spectral analysis of a Boolean function is proposed for approximating the decision boundary of an ensemble of classifiers, and an intuitive explanation of computing Walsh coefficients for the functional approximation is provided. It is shown that the difference between the first- and third-order coefficient approximations is a good indicator of optimal base classifier complexity. When combining neural networks, the experimental results on a variety of artificial and real two-class problems demonstrate under what circumstances ensemble performance can be improved. For tuned base classifiers, the first-order coefficients provide performance similar to the majority vote. However, for weak/fast base classifiers, higher order coefficient approximation may give better performance. It is also shown that higher order coefficient approximation is superior to the Adaboost logarithmic weighting rule when boosting weak decision tree base classifiers.

Sign Language Recognition (SLR) has been an active research field for the last two decades. However, most research to date has considered SLR as a naive gesture recognition problem. SLR seeks to recognize a sequence of continuous signs but neglects the underlying rich grammatical and linguistic structures of sign language that differ from spoken language. In contrast, we introduce the Sign Language Translation (SLT) problem. Here, the objective is to generate spoken language translations from sign language videos, taking into account the different word orders and grammar. We formalize SLT in the framework of Neural Machine Translation (NMT) for both end-to-end and pretrained settings (using expert knowledge). This allows us to jointly learn the spatial representations, the underlying language model, and the mapping between sign and spoken language. To evaluate the performance of Neural SLT, we collected the first publicly available Continuous SLT dataset, RWTH-PHOENIX-Weather 2014T1. It provides spoken language translations and gloss level annotations for German Sign Language videos of weather broadcasts. Our dataset contains over .95M frames with >67K signs from a sign vocabulary of >1K and >99K words from a German vocabulary of >2.8K. We report quantitative and qualitative results for various SLT setups to underpin future research in this newly established field. The upper bound for translation performance is calculated at 19.26 BLEU-4, while our end-to-end frame-level and gloss-level tokenization networks were able to achieve 9.58 and 18.13 respectively.

We propose a novel deep learning approach to solve simultaneous alignment and recognition problems (referred to as “Sequence-to-sequence” learning). We decompose the problem into a series of specialised expert systems referred to as SubUNets. The spatio-temporal relationships between these SubUNets are then modelled to solve the task, while remaining trainable end-to-end. The approach mimics human learning and educational techniques, and has a number of significant advantages. SubUNets allow us to inject domain-specific expert knowledge into the system regarding suitable intermediate representations. They also allow us to implicitly perform transfer learning between different interrelated tasks, which also allows us to exploit a wider range of more varied data sources. In our experiments we demonstrate that each of these properties serves to significantly improve the performance of the overarching recognition system, by better constraining the learning problem. The proposed techniques are demonstrated in the challenging domain of sign language recognition. We demonstrate state-of-the-art performance on hand-shape recognition (outperforming previous techniques by more than 30%). Furthermore, we are able to obtain comparable sign recognition rates to previous research, without the need for an alignment step to segment out the signs for recognition.

In this paper, we propose using 3D Convolutional Neural Networks for large scale user-independent continuous gesture recognition. We have trained an end-to-end deep network for continuous gesture recognition (jointly learning both the feature representation and the classifier). The network performs three-dimensional (i.e. space-time) convolutions to extract features related to both the appearance and motion from volumes of color frames. Space-time invariance of the extracted features is encoded via pooling layers. The earlier stages of the network are partially initialized using the work of Tran et al. before being adapted to the task of gesture recognition. An earlier version of the proposed method, which was trained for 11,250 iterations, was submitted to ChaLearn 2016 Continuous Gesture Recognition Challenge and ranked 2nd with the Mean Jaccard Index Score of 0.269235. When the proposed method was further trained for 28,750 iterations, it achieved state-of-the-art performance on the same dataset, yielding a 0.314779 Mean Jaccard Index Score.