Jonas Beskow

Simon Alexanderson, Rajmund Nagy, Jonas Beskow et al.

Diffusion models have experienced a surge of interest as highly expressive yet efficiently trainable probabilistic models. We show that these models are an excellent fit for synthesising human motion that co-occurs with audio, e.g., dancing and co-speech gesticulation, since motion is complex and highly ambiguous given audio, calling for a probabilistic description. Specifically, we adapt the DiffWave architecture to model 3D pose sequences, putting Conformers in place of dilated convolutions for improved modelling power. We also demonstrate control over motion style, using classifier-free guidance to adjust the strength of the stylistic expression. Experiments on gesture and dance generation confirm that the proposed method achieves top-of-the-line motion quality, with distinctive styles whose expression can be made more or less pronounced. We also synthesise path-driven locomotion using the same model architecture. Finally, we generalise the guidance procedure to obtain product-of-expert ensembles of diffusion models and demonstrate how these may be used for, e.g., style interpolation, a contribution we believe is of independent interest. See https://www.speech.kth.se/research/listen-denoise-action/ for video examples, data, and code.

Shivam Mehta, Ruibo Tu, Jonas Beskow et al.

We introduce Matcha-TTS, a new encoder-decoder architecture for speedy TTS acoustic modelling, trained using optimal-transport conditional flow matching (OT-CFM). This yields an ODE-based decoder capable of high output quality in fewer synthesis steps than models trained using score matching. Careful design choices additionally ensure each synthesis step is fast to run. The method is probabilistic, non-autoregressive, and learns to speak from scratch without external alignments. Compared to strong pre-trained baseline models, the Matcha-TTS system has the smallest memory footprint, rivals the speed of the fastest models on long utterances, and attains the highest mean opinion score in a listening test. Please see https://shivammehta25.github.io/Matcha-TTS/ for audio examples, code, and pre-trained models.

Anna Deichler, Shivam Mehta, Simon Alexanderson et al.

This paper describes a system developed for the GENEA (Generation and Evaluation of Non-verbal Behaviour for Embodied Agents) Challenge 2023. Our solution builds on an existing diffusion-based motion synthesis model. We propose a contrastive speech and motion pretraining (CSMP) module, which learns a joint embedding for speech and gesture with the aim to learn a semantic coupling between these modalities. The output of the CSMP module is used as a conditioning signal in the diffusion-based gesture synthesis model in order to achieve semantically-aware co-speech gesture generation. Our entry achieved highest human-likeness and highest speech appropriateness rating among the submitted entries. This indicates that our system is a promising approach to achieve human-like co-speech gestures in agents that carry semantic meaning.

Shivam Mehta, Ambika Kirkland, Harm Lameris et al.

Neural HMMs are a type of neural transducer recently proposed for sequence-to-sequence modelling in text-to-speech. They combine the best features of classic statistical speech synthesis and modern neural TTS, requiring less data and fewer training updates, and are less prone to gibberish output caused by neural attention failures. In this paper, we combine neural HMM TTS with normalising flows for describing the highly non-Gaussian distribution of speech acoustics. The result is a powerful, fully probabilistic model of durations and acoustics that can be trained using exact maximum likelihood. Experiments show that a system based on our proposal needs fewer updates than comparable methods to produce accurate pronunciations and a subjective speech quality close to natural speech. Please see https://shivammehta25.github.io/OverFlow/ for audio examples and code.

Shivam Mehta, Siyang Wang, Simon Alexanderson et al.

With read-aloud speech synthesis achieving high naturalness scores, there is a growing research interest in synthesising spontaneous speech. However, human spontaneous face-to-face conversation has both spoken and non-verbal aspects (here, co-speech gestures). Only recently has research begun to explore the benefits of jointly synthesising these two modalities in a single system. The previous state of the art used non-probabilistic methods, which fail to capture the variability of human speech and motion, and risk producing oversmoothing artefacts and sub-optimal synthesis quality. We present the first diffusion-based probabilistic model, called Diff-TTSG, that jointly learns to synthesise speech and gestures together. Our method can be trained on small datasets from scratch. Furthermore, we describe a set of careful uni- and multi-modal subjective tests for evaluating integrated speech and gesture synthesis systems, and use them to validate our proposed approach. Please see https://shivammehta25.github.io/Diff-TTSG/ for video examples, data, and code.

Shivam Mehta, Ruibo Tu, Simon Alexanderson et al.

As text-to-speech technologies achieve remarkable naturalness in read-aloud tasks, there is growing interest in multimodal synthesis of verbal and non-verbal communicative behaviour, such as spontaneous speech and associated body gestures. This paper presents a novel, unified architecture for jointly synthesising speech acoustics and skeleton-based 3D gesture motion from text, trained using optimal-transport conditional flow matching (OT-CFM). The proposed architecture is simpler than the previous state of the art, has a smaller memory footprint, and can capture the joint distribution of speech and gestures, generating both modalities together in one single process. The new training regime, meanwhile, enables better synthesis quality in much fewer steps (network evaluations) than before. Uni- and multimodal subjective tests demonstrate improved speech naturalness, gesture human-likeness, and cross-modal appropriateness compared to existing benchmarks. Please see https://shivammehta25.github.io/Match-TTSG/ for video examples and code.

Anna Deichler, Jim O'Regan, Fethiye Irmak Dogan et al.

Grounding language in the physical world requires AI systems to interpret references that emerge dynamically during conversation. While current vision-language models (VLMs) excel at static image tasks, they struggle to resolve ambiguous expressions in spontaneous, multi-turn dialogue. We address this gap by introducing (1) a benchmark for referential communication in dynamic 3D environments, built from 6.7 hours of egocentric VR interaction with synchronized speech, motion, gaze, and 3D scene geometry, and (2) a two-stage grounding pipeline that explicitly resolves conversational ambiguity before visual localization. The benchmark includes over 4,200 manually verified referring expressions spanning full, partitive, and pronominal types. Our contextual rewriting approach improves grounding performance by 11-22 percentage points on average, with a pure detector (GroundingDINO) reaching 56.7% on pronominals after rewriting, nearly double the best end-to-end baseline. Results demonstrate that decoupling linguistic reasoning from visual perception is more effective than end-to-end approaches for conversational grounding.

Anna Deichler, Simon Alexanderson, Jonas Beskow

This paper focuses on enhancing human-agent communication by integrating spatial context into virtual agents' non-verbal behaviors, specifically gestures. Recent advances in co-speech gesture generation have primarily utilized data-driven methods, which create natural motion but limit the scope of gestures to those performed in a void. Our work aims to extend these methods by enabling generative models to incorporate scene information into speech-driven gesture synthesis. We introduce a novel synthetic gesture dataset tailored for this purpose. This development represents a critical step toward creating embodied conversational agents that interact more naturally with their environment and users.

Anna Deichler, Jim O'Regan, Jonas Beskow

In this paper, we present a novel dataset captured using a VR headset to record conversations between participants within a physics simulator (AI2-THOR). Our primary objective is to extend the field of co-speech gesture generation by incorporating rich contextual information within referential settings. Participants engaged in various conversational scenarios, all based on referential communication tasks. The dataset provides a rich set of multimodal recordings such as motion capture, speech, gaze, and scene graphs. This comprehensive dataset aims to enhance the understanding and development of gesture generation models in 3D scenes by providing diverse and contextually rich data.

Taras Kucherenko, Jonas Beskow, Hedvig Kjellström

Optical motion capture systems have become a widely used technology in various fields, such as augmented reality, robotics, movie production, etc. Such systems use a large number of cameras to triangulate the position of optical markers.The marker positions are estimated with high accuracy. However, especially when tracking articulated bodies, a fraction of the markers in each timestep is missing from the reconstruction. In this paper, we propose to use a neural network approach to learn how human motion is temporally and spatially correlated, and reconstruct missing markers positions through this model. We experiment with two different models, one LSTM-based and one time-window-based. Both methods produce state-of-the-art results, while working online, as opposed to most of the alternative methods, which require the complete sequence to be known. The implementation is publicly available at https://github.com/Svito-zar/NN-for-Missing-Marker-Reconstruction .

Anna Deichler, Siyang Wang, Simon Alexanderson et al.

One of the main goals of robotics and intelligent agent research is to enable natural communication with humans in physically situated settings. While recent work has focused on verbal modes such as language and speech, non-verbal communication is crucial for flexible interaction. We present a framework for generating pointing gestures in embodied agents by combining imitation and reinforcement learning. Using a small motion capture dataset, our method learns a motor control policy that produces physically valid, naturalistic gestures with high referential accuracy. We evaluate the approach against supervised learning and retrieval baselines in both objective metrics and a virtual reality referential game with human users. Results show that our system achieves higher naturalness and accuracy than state-of-the-art supervised models, highlighting the promise of imitation-RL for communicative gesture generation and its potential application to robots.

Anna Deichler, Siyang Wang, Simon Alexanderson et al.

Pointing is a key mode of interaction with robots, yet most prior work has focused on recognition rather than generation. We present a motion capture dataset of human pointing gestures covering diverse styles, handedness, and spatial targets. Using reinforcement learning with motion imitation, we train policies that reproduce human-like pointing while maximizing precision. Results show our approach enables context-aware pointing behaviors in simulation, balancing task performance with natural dynamics.

Shivam Mehta, Anna Deichler, Jim O'Regan et al.

Although humans engaged in face-to-face conversation simultaneously communicate both verbally and non-verbally, methods for joint and unified synthesis of speech audio and co-speech 3D gesture motion from text are a new and emerging field. These technologies hold great promise for more human-like, efficient, expressive, and robust synthetic communication, but are currently held back by the lack of suitably large datasets, as existing methods are trained on parallel data from all constituent modalities. Inspired by student-teacher methods, we propose a straightforward solution to the data shortage, by simply synthesising additional training material. Specifically, we use unimodal synthesis models trained on large datasets to create multimodal (but synthetic) parallel training data, and then pre-train a joint synthesis model on that material. In addition, we propose a new synthesis architecture that adds better and more controllable prosody modelling to the state-of-the-art method in the field. Our results confirm that pre-training on large amounts of synthetic data improves the quality of both the speech and the motion synthesised by the multimodal model, with the proposed architecture yielding further benefits when pre-trained on the synthetic data. See https://shivammehta25.github.io/MAGI/ for example output.

Axel Wiebe Werner, Jonas Beskow, Anna Deichler

Gestures are central to human communication, enriching interactions through non-verbal expression. Virtual avatars increasingly use AI-generated gestures to enhance life-likeness, yet evaluations have largely been confined to 2D. Virtual Reality (VR) provides an immersive alternative that may affect how gestures are perceived. This paper presents a comparative evaluation of computer-generated gestures in VR and 2D, examining three models from the 2023 GENEA Challenge. Results show that gestures viewed in VR were rated slightly higher on average, with the strongest effect observed for motion-capture "true movement." While model rankings remained consistent across settings, VR influenced participants' overall perception and offered unique benefits over traditional 2D evaluation.

Téo Guichoux, Théodor Lemerle, Shivam Mehta et al.

Human communication is multimodal, with speech and gestures tightly coupled, yet most computational methods for generating speech and gestures synthesize them sequentially, weakening synchrony and prosody alignment. We introduce Gelina, a unified framework that jointly synthesizes speech and co-speech gestures from text using interleaved token sequences in a discrete autoregressive backbone, with modality-specific decoders. Gelina supports multi-speaker and multi-style cloning and enables gesture-only synthesis from speech inputs. Subjective and objective evaluations demonstrate competitive speech quality and improved gesture generation over unimodal baselines.

Anna Deichler, Jim O'Regan, Teo Guichoux et al.

Human motion generation has advanced rapidly in recent years, yet the critical problem of creating spatially grounded, context-aware gestures has been largely overlooked. Existing models typically specialize either in descriptive motion generation, such as locomotion and object interaction, or in isolated co-speech gesture synthesis aligned with utterance semantics. However, both lines of work often treat motion and environmental grounding separately, limiting advances toward embodied, communicative agents. To address this gap, our work introduces a multimodal dataset and framework for grounded gesture generation, combining two key resources: (1) a synthetic dataset of spatially grounded referential gestures, and (2) MM-Conv, a VR-based dataset capturing two-party dialogues. Together, they provide over 7.7 hours of synchronized motion, speech, and 3D scene information, standardized in the HumanML3D format. Our framework further connects to a physics-based simulator, enabling synthetic data generation and situated evaluation. By bridging gesture modeling and spatial grounding, our contribution establishes a foundation for advancing research in situated gesture generation and grounded multimodal interaction. Project page: https://groundedgestures.github.io/

Patrik Jonell, Anna Deichler, Ilaria Torre et al.

In this paper we present a pilot study which investigates how non-verbal behavior affects social influence in social robots. We also present a modular system which is capable of controlling the non-verbal behavior based on the interlocutor's facial gestures (head movements and facial expressions) in real time, and a study investigating whether three different strategies for facial gestures ("still", "natural movement", i.e. movements recorded from another conversation, and "copy", i.e. mimicking the user with a four second delay) has any affect on social influence and decision making in a "survival task". Our preliminary results show there was no significant difference between the three conditions, but this might be due to among other things a low number of study participants (12).

Shivam Mehta, Éva Székely, Jonas Beskow et al.

Neural sequence-to-sequence TTS has achieved significantly better output quality than statistical speech synthesis using HMMs. However, neural TTS is generally not probabilistic and uses non-monotonic attention. Attention failures increase training time and can make synthesis babble incoherently. This paper describes how the old and new paradigms can be combined to obtain the advantages of both worlds, by replacing attention in neural TTS with an autoregressive left-right no-skip hidden Markov model defined by a neural network. Based on this proposal, we modify Tacotron 2 to obtain an HMM-based neural TTS model with monotonic alignment, trained to maximise the full sequence likelihood without approximation. We also describe how to combine ideas from classical and contemporary TTS for best results. The resulting example system is smaller and simpler than Tacotron 2, and learns to speak with fewer iterations and less data, whilst achieving comparable naturalness prior to the post-net. Our approach also allows easy control over speaking rate.

Guillermo Valle-Pérez, Gustav Eje Henter, Jonas Beskow et al.

Dance requires skillful composition of complex movements that follow rhythmic, tonal and timbral features of music. Formally, generating dance conditioned on a piece of music can be expressed as a problem of modelling a high-dimensional continuous motion signal, conditioned on an audio signal. In this work we make two contributions to tackle this problem. First, we present a novel probabilistic autoregressive architecture that models the distribution over future poses with a normalizing flow conditioned on previous poses as well as music context, using a multimodal transformer encoder. Second, we introduce the currently largest 3D dance-motion dataset, obtained with a variety of motion-capture technologies, and including both professional and casual dancers. Using this dataset, we compare our new model against two baselines, via objective metrics and a user study, and show that both the ability to model a probability distribution, as well as being able to attend over a large motion and music context are necessary to produce interesting, diverse, and realistic dance that matches the music.

Patrik Jonell, Taras Kucherenko, Ilaria Torre et al.

Conducting user studies is a crucial component in many scientific fields. While some studies require participants to be physically present, other studies can be conducted both physically (e.g. in-lab) and online (e.g. via crowdsourcing). Inviting participants to the lab can be a time-consuming and logistically difficult endeavor, not to mention that sometimes research groups might not be able to run in-lab experiments, because of, for example, a pandemic. Crowdsourcing platforms such as Amazon Mechanical Turk (AMT) or Prolific can therefore be a suitable alternative to run certain experiments, such as evaluating virtual agents. Although previous studies investigated the use of crowdsourcing platforms for running experiments, there is still uncertainty as to whether the results are reliable for perceptual studies. Here we replicate a previous experiment where participants evaluated a gesture generation model for virtual agents. The experiment is conducted across three participant pools -- in-lab, Prolific, and AMT -- having similar demographics across the in-lab participants and the Prolific platform. Our results show no difference between the three participant pools in regards to their evaluations of the gesture generation models and their reliability scores. The results indicate that online platforms can successfully be used for perceptual evaluations of this kind.

Andreas Wedenborn, Preben Wik, Olov Engwall et al.

This study investigates the effect of a physical robot taking the role of a teacher or exercise partner in a language learning exercise. In order to investigate this, an application was developed enabling a 2:nd language learning vocabulary exercise in three different conditions. In the first condition the learner would receive tutoring from a disembodied voice, in the second condition the tutor would be embodied by an animated avatar on a computer screen, and in the final condition the tutor was a physical robotic head with a 3D animated face mask. A Russian language vocabulary exercise with 15 subjects was conducted. None of the subjects reported any Russian language skills prior to the exercises. Each subject were taught a set of 9 words in each of the three conditions during a practice phase, and were then asked to recall the words in a test phase. Results show that the recall of the words practiced with the physical robot were significantly higher than that of the words practiced with the avatar on the screen or with the disembodied voice.

Kalin Stefanov, Jonas Beskow, Giampiero Salvi

This paper presents a self-supervised method for visual detection of the active speaker in a multi-person spoken interaction scenario. Active speaker detection is a fundamental prerequisite for any artificial cognitive system attempting to acquire language in social settings. The proposed method is intended to complement the acoustic detection of the active speaker, thus improving the system robustness in noisy conditions. The method can detect an arbitrary number of possibly overlapping active speakers based exclusively on visual information about their face. Furthermore, the method does not rely on external annotations, thus complying with cognitive development. Instead, the method uses information from the auditory modality to support learning in the visual domain. This paper reports an extensive evaluation of the proposed method using a large multi-person face-to-face interaction dataset. The results show good performance in a speaker dependent setting. However, in a speaker independent setting the proposed method yields a significantly lower performance. We believe that the proposed method represents an essential component of any artificial cognitive system or robotic platform engaging in social interactions.

Patrik Jonell, Joseph Mendelson, Thomas Storskog et al.

This paper presents the EACare project, an ambitious multi-disciplinary collaboration with the aim to develop an embodied system, capable of carrying out neuropsychological tests to detect early signs of dementia, e.g., due to Alzheimer's disease. The system will use methods from Machine Learning and Social Robotics, and be trained with examples of recorded clinician-patient interactions. The interaction will be developed using a participatory design approach. We describe the scope and method of the project, and report on a first Wizard of Oz prototype.