Abstract Motion tracking has been an important technique for imitating human‐like movement from large‐scale datasets in physics‐based motion synthesis. However, existing approaches focus on tracking either single character or a particular type of interaction, limiting their ability to handle contact‐rich interactions. Extending single‐character tracking approaches suffers from the instability due to the challenge of forces transferred through contacts. Contact‐rich interactions requires levels of control, which places much greater demands on model capacity. To this end, we propose a robust tracking method based on progressive neural network (PNN) where multiple experts are specialized in learning skills of various difficulties. Our method learns to assign training samples to experts automatically without requiring manually scheduling. Both qualitative and quantitative results show that our method delivers more stable motion tracking in densely interactive movements while enabling more efficient model training.

Diffusion models are learning pattern-learning systems to model and sample from data distributions with three functional components namely the forward process, the reverse process, and the sampling process. The components of diffusion models have gained significant attention with many design factors being considered in common practice. Existing reviews have primarily focused on higher-level solutions, covering less on the design fundamentals of components. This study seeks to address this gap by providing a comprehensive and coherent review of seminal designable factors within each functional component of diffusion models. This provides a finer-grained perspective of diffusion models, benefiting future studies in the analysis of individual components, the design factors for different purposes, and the implementation of diffusion models. • A literature review of design fundamentals of diffusion models. • Identified the three functional components of diffusion models. • Examined the key factors and their functionalities of the components. • Discussed the popular designs and seminal works for each factor.

Battery-free tags offer a power-efficient solution for the wireless connection in Internet of Things (IoT), where the backscatter communication that is compatible with widely-deployed protocols such as Bluetooth Low Energy (BLE) significantly reduces the hardware cost thanks to the seamless integration into the existing infrastructure. However, there are three main shortcomings in the existing battery-free BLE tags: 1)The tag-to-access point (AP, uplink) communication ranges are limited to 97 meters at -10dBm incident power, which are not long enough for some outdoor scenes. 2) The AP-to-tag (downlink) communication in the state of the arts has not exceeded a 1Mbps data rate and a 4m range, disabling remote tag configuration. 3)The tag-to-tag communication has not been realized in a battery-free way, which cannot construct a passive IoT network. In this work, we demonstrate an integrated bidirectional passive BLE chip that conducts battery-free BLE communication in tag-to-tablet/smartphone, tablet/smartphone-to-tag, and tag-to-tag modes. The chip is implemented in a 65nm CMOS process. An all-digital intermediate frequency (IF) shaping technique is proposed to extend the tag-to-tablet/smartphone (uplink) communication range to 160m at$7.16\mu $W power consumption and a -10dBm incident tone. In addition, a$2^{nd}$-order intermodulation (IM2) based charge-domain GFSK demodulation technique is proposed to enable 1Mbps 12m downlink communication at$6.8\mu $W power consumption, which realizes both tablet/smartphone-to-tag and tag-to-tag communication in a fully battery-free way. As a result, the bidirectional passive BLE chip offers a potential solution for future battery-free IoT mesh network.

Motion in-betweening is the problem to synthesize movement between keyposes. Traditional research focused primarily on single characters. Extending them to densely interacting characters is highly challenging, as it demands precise spatial-temporal correspondence between the characters to maintain the interaction, while creating natural transitions towards predefined keyposes. In this research, we present a method for long-horizon interaction in-betweening that enables two characters to engage and respond to one another naturally. To effectively represent and synthesize interactions, we propose a novel solution called Cross-Space In-Betweening, which models the interactions of each character across different conditioning representation spaces. We further observe that the significantly increased constraints in interacting characters heavily limit the solution space, leading to degraded motion quality and diminished interaction over time. To enable long-horizon synthesis, we present two solutions to maintain long-term interaction and motion quality, thereby keeping synthesis in the stable region of the solution space. We first sustain interaction quality by identifying periodic interaction patterns through adversarial learning. We further maintain the motion quality by learning to refine the drifted latent space and prevent pose error accumulation. We demonstrate that our approach produces realistic, controllable, and long-horizon in-between motions of two characters with dynamic boxing and dancing actions across multiple keyposes, supported by extensive quantitative evaluations and user studies.

Abstract We propose a real‐time method for reactive motion synthesis based on the known trajectory of input character, predicting instant reactions using only historical, user‐controlled motions. Our method handles the uncertainty of future movements by introducing an intention predictor, which forecasts key joint intentions to make pose prediction more deterministic from the historical interaction. The intention is later encoded into the latent space of its reactive motion, matched with a codebook which represents mappings between input and output. It samples from the categorical distribution for pose generation and strengthens model robustness through adversarial training. Unlike previous offline approaches, the system can recursively generate intentions and reactive motions using feedback from earlier steps, enabling real‐time, long‐term realistic interactive synthesis. Both quantitative and qualitative experiments show our approach outperforms other matching‐based motion synthesis approaches, delivering superior stability and generalisability. In our method, user can also actively influence the outcome by controlling the moving directions, creating a personalised interaction path that deviates from predefined trajectories.

Recently, methods for skeleton-based human activity recognition have been shown to be vulnerable to adversarial attacks. However, these attack methods require either the full knowledge of the victim (i.e. white-box attacks), access to training data (i.e. transfer-based attacks) or frequent model queries (i.e. black-box attacks). All their requirements are highly restrictive, raising the question of how detrimental the vulnerability is. In this paper, we show that the vulnerability indeed exists. To this end, we consider a new attack task: the attacker has no access to the victim model or the training data or labels, where we coin the term hard no-box attack. Specifically, we first learn a motion manifold where we define an adversarial loss to compute a new gradient for the attack, named skeleton-motioninformed (SMI) gradient. Our gradient contains information of the motion dynamics, which is different from existing gradient-based attack methods that compute the loss gradient assuming each dimension in the data is independent. The SMI gradient can augment many gradient-based attack methods, leading to a new family of no-box attack methods. Extensive evaluation and comparison show that our method imposes a real threat to existing classifiers. They also show that the SMI gradient improves the transferability and imperceptibility of adversarial samples in both no-box and transfer-based black-box settings.

Acquiring the virtual equivalent of exhibits, such as sculptures, in virtual reality (VR) museums, can be labour-intensive and sometimes infeasible. Deep learning based 3D reconstruction approaches allow us to recover 3D shapes from 2D observations, among which single-view-based approaches can reduce the need for human intervention and specialised equipment in acquiring 3D sculptures for VR museums. However, there exist two challenges when attempting to use the well-researched human reconstruction methods: limited data availability and domain shift. Considering sculptures are usually related to humans, we propose our unsupervised 3D domain adaptation method for adapting a single-view 3D implicit reconstruction model from the source (real-world humans) to the target (sculptures) domain. We have compared the generated shapes with other methods and conducted ablation studies as well as a user study to demonstrate the effectiveness of our adaptation method. We also deploy our results in a VR application.

The size of wireless systems is required to be reduced in many applications, such as ultra-low-power sensor nodes and wearable/implantable devices, where battery and crystal are the two main bottlenecks in system miniaturization. In recent years, battery-free radios based on wireless power transfer (WPT) have shown great potential in miniature wireless systems, while a reliable on-chip clock without a crystal remains a design challenge. Conventional methods utilized the RF WPT tone as the reference for clock generation, but the high RF frequency leads to high power consumption. In comparison, using a lower WPT frequency results in an antenna with a larger size. In this work, the$2^{\mathrm{nd}}$-order inter-modulation (IM2) component of the two RF WPT tones is extracted to lock an on-chip oscillator, providing a low-jitter PVT-robust clock. In this way, the wireless systems can benefit from: 1) The clock recovery circuits operate at a low IM2 frequency, reducing the power consumption. 2) The WPT can be set to a high RF frequency to minimize the antenna. Fabricated in 65 nm CMOS process, the proposed crystal-less clock generator takes a small area of 0.023 mm2 in a wireless system chip. Measured results show −92 dBc/Hz@10 kHz phase noise and 6.8$\mu \text{W}$power.