Inspired by the Transformer's adaptive capability in natural language processing and computer vision domains, and supported by CLIP's strong semantic prior knowledge, we propose a novel generative model named STMotionFormer based on improved Transformer block and CLIP for language-guided 3D skeleton-based human motion synthesis. The skeleton sequence, as a kind of spatial temporal dynamic data, not only contains temporal information but also consists of a natural graph structure by different body parts. We focus on this, propose Temporal Transformer block (T-Former) applies attention mechanism along the temporal dimension to capture the long-term relationship, and propose spatial Transformer block (S-Former) applies attention mechanism in the spatial dimension to aggregate and update joints' spatial features. Therefore, our approach can automatically explore both spatial and temporal information, which can generate skeleton sequence with reasonable graph structure as well as global time relationship. Moreover, powerful semantic prior knowledge of CLIP are injected into our motion-language joint manifold. We evaluate our method on the public KIT motion language dataset that contains manually annotated 3D pose sequences. Experimental results show that our model outperforms the state-of-the-art in terms of APE and AVE respectively. Qualitative visualization results indicate that our model can generate human motions that are more compatible with semantic information even than the GroundTruth.

Opinion dynamics, which concerns how opinions evolve and spread in social networks has been widely studied during past years, and a lot of classical models have been proposed to describe the opinion diffusion process. However, most existing leader-follower-relationship based models ignore the influence of normal individuals and do not consider the feedback effect of opinion difference on individuals, which are important for opinion spread. In this paper, inspired by two well-known social theories: Emotional Mobilization and Social Judgement Theory, we first propose a method to identify individual influence factor based on both network topology and personal behavior attributes. Then, we propose a novel opinion evolution model named Dynamic Adaptive Individual Weighting model which focuses on individual heterogeneity and considers the opinion difference assimilation effect. In this model, the influence weight of an agent's neighbour on the agent can be dynamically affected by their opinion difference and adaptively adjusted based on the neighbour's relative influence factor. Moreover, environmental noise is also introduced to assimilate realistic situations' uncertainty. Experimental results on 12 real and 2 generated network datasets show that our proposed model can precisely reflect the evolution process and trend of opinions over different social networks. The study can enable decision-makers better understand the fundamental processes of opinion diffusion and design more efficient strategies for political or business activities.