Synthesizing human motion through learning techniques is becoming an increasingly popular approach to alleviating the requirement of new data capture to produce animations. Learning to move naturally from music, i.e., to dance, is one of the more complex motions humans often perform effortlessly. Each dance movement is unique, yet such movements maintain the core characteristics of the dance style. Most approaches addressing this problem with classical convolutional and recursive neural models undergo training and variability issues due to the non-Euclidean geometry of the motion manifold structure. In this paper, we design a novel method based on graph convolutional networks to tackle the problem of automatic dance generation from audio information. Our method uses an adversarial learning scheme conditioned on the input music audios to create natural motions preserving the key movements of different music styles. We evaluate our method with three quantitative metrics of generative methods and a user study. The results suggest that the proposed GCN model outperforms the state-of-the-art dance generation method conditioned on music in different experiments. Moreover, our graph-convolutional approach is simpler, easier to be trained, and capable of generating more realistic motion styles regarding qualitative and different quantitative metrics. It also presented a visual movement perceptual quality comparable to real motion data. The dataset and project are publicly available at: https://www.verlab.dcc. ufmg.br/motion-analysis/cag2020.
The performance of RBF neural networks depends on the choice of the parameters of the radial base functions that are usually defined by the user or by validation methods. In this work geometric information of the structure of the input data and a linear discriminant are used to select the neurons of the hidden layer. The proposed approach showed promising results being able to reduce the number of neurons of the hidden layer compared to other methods, thus also reducing the complexity of the solutions. Resumo: O desempenho de redes neurais RBF depende da escolha dos parâmetros das funções de bases radiais que são normalmente definidas pelo usuário ou por métodos de validação. Neste trabalho são usadas informações geométricas da estrutura dos dados de entrada e um discriminante linear para selecionar os neurônios da camada escondida. A abordagem proposta mostrou resultados promissores sendo capaz de reduzir o número de neurônios da camada escondida comparada a outros métodos, assim reduzindo também a complexidade das soluções.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.