20,948 Hits in 4.0 sec

Learning to Generate Diverse Dance Motions with Transformer [article]

Jiaman Li, Yihang Yin, Hang Chu, Yi Zhou, Tingwu Wang, Sanja Fidler, Hao Li
2020 arXiv   pre-print
We also present a novel two-stream motion transformer generative model, which can generate motion sequences with high flexibility.  ...  We introduce a complete system for dance motion synthesis, which can generate complex and highly diverse dance sequences given an input music sequence.  ...  In this work, we introduce a novel system that can synthesize diverse dance motions by learning from a large-scale dataset with a comprehensive set of highly diverse dance movements.  ... 
arXiv:2008.08171v1 fatcat:kp57scjce5cnlejdqpukvhurem

Transflower: probabilistic autoregressive dance generation with multimodal attention [article]

Guillermo Valle-Pérez, Gustav Eje Henter, Jonas Beskow, André Holzapfel, Pierre-Yves Oudeyer, Simon Alexanderson
2021 arXiv   pre-print
attend over a large motion and music context are necessary to produce interesting, diverse, and realistic dance that matches the music.  ...  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.  ...  Learning to Generate Diverse Dance Motions with Transformer. arXiv preprint Heewoo Jun, Tom B Brown, Prafulla Dhariwal, Scott Gray, et al. 2020.  ... 
arXiv:2106.13871v1 fatcat:ofkengygabcfdlbcwsl6vk4rmq

AI Choreographer: Music Conditioned 3D Dance Generation with AIST++ [article]

Ruilong Li, Shan Yang, David A. Ross, Angjoo Kanazawa
2021 arXiv   pre-print
We present AIST++, a new multi-modal dataset of 3D dance motion and music, along with FACT, a Full-Attention Cross-modal Transformer network for generating 3D dance motion conditioned on music.  ...  We show that naively applying sequence models such as transformers to this dataset for the task of music conditioned 3D motion generation does not produce satisfactory 3D motion that is well correlated  ...  Generation Diversity We also evaluate our model's ability to generate diverse dance motions when given various input music compared with the baseline methods.  ... 
arXiv:2101.08779v3 fatcat:gsdbkgtq7vb7xhxsfrmfb43tke

Bailando: 3D Dance Generation by Actor-Critic GPT with Choreographic Memory [article]

Li Siyao, Weijiang Yu, Tianpei Gu, Chunze Lin, Quan Wang, Chen Qian, Chen Change Loy, Ziwei Liu
2022 arXiv   pre-print
To achieve synchronized alignment between diverse motion tempos and music beats, we introduce an actor-critic-based reinforcement learning scheme to the GPT with a newly-designed beat-align reward function  ...  sequence to a quantized codebook, 2) an actor-critic Generative Pre-trained Transformer (GPT) that composes these units to a fluent dance coherent to the music.  ...  This research is conducted in collaboration with SenseTime. This work is supported by NTU NAP and A*STAR through the Industry Alignment Fund -Industry Collaboration Projects Grant.  ... 
arXiv:2203.13055v2 fatcat:rw5ixqeyynevdcsdkcjs7stfka

Dance Generation with Style Embedding: Learning and Transferring Latent Representations of Dance Styles [article]

Xinjian Zhang, Yi Xu, Su Yang, Longwen Gao, Huyang Sun
2021 arXiv   pre-print
Choreography refers to creation of dance steps and motions for dances according to the latent knowledge in human mind, where the created dance motions are in general style-specific and consistent.  ...  So far, such latent style-specific knowledge about dance styles cannot be represented explicitly in human language and has not yet been learned in previous works on music-to-dance generation tasks.  ...  As descripted in Figure 2 , we apply a set of transformer music encoders with a Bi-LSTM dance generator in music-to-dance generator.  ... 
arXiv:2104.14802v1 fatcat:jkecjargvrc2naypu4huynngra

DanceConv: Dance Motion Generation with Convolutional Networks

Kosmas Kritsis, Aggelos Gkiokas, Aggelos Pikrakis, Vassilis Katsouros
2022 IEEE Access  
Subjective results confirm the effectiveness of our approach to synthesize diverse dance motion sequences, reporting a 6% increase in style consistency preference according to 1800 answers provided by  ...  Based on this outcome, we train the proposed multimodal architecture with two different approaches, namely teacher-forcing and self-supervised curriculum learning, to deal with the autoregressive error  ...  They also would like to thank all the colleagues and friends that participated in both user-studies and provided their valuable feedback.  ... 
doi:10.1109/access.2022.3169782 fatcat:abjqqrrww5bulh5tadggkwzk4u

DanceFormer: Music Conditioned 3D Dance Generation with Parametric Motion Transformer [article]

Buyu Li, Yongchi Zhao, Zhelun Shi, Lu Sheng
2021 arXiv   pre-print
In this paper, we reformulate it by a two-stage process, ie, a key pose generation and then an in-between parametric motion curve prediction, where the key poses are easier to be synchronized with the  ...  Previous works treat this task as sequence generation, however, it is challenging to render a music-aligned long-term sequence with high kinematic complexity and coherent movements.  ...  The generation diversity is ensured by the adversarial learning scheme.  ... 
arXiv:2103.10206v4 fatcat:tqsqxlrhffaxxaujfsgake5zwm

Dance Revolution: Long-Term Dance Generation with Music via Curriculum Learning [article]

Ruozi Huang, Huang Hu, Wei Wu, Kei Sawada, Mi Zhang, Daxin Jiang
2021 arXiv   pre-print
In this paper, we formalize the music-conditioned dance generation as a sequence-to-sequence learning problem and devise a novel seq2seq architecture to efficiently process long sequences of music features  ...  Furthermore, we propose a novel curriculum learning strategy to alleviate error accumulation of autoregressive models in long motion sequence generation, which gently changes the training process from  ...  generation with music. (2) We propose a novel curriculum learning strategy to alleviate error accumulation of autoregressive models in long motion sequence generation. (3) To facilitate long-term dance  ... 
arXiv:2006.06119v7 fatcat:gmii276zgjdkxcisfsp4sb5ubm

Dual Learning Music Composition and Dance Choreography [article]

Shuang Wu, Zhenguang Li, Shijian Lu, Li Cheng
2022 arXiv   pre-print
Experimental results demonstrate that our dual learning framework improves individual task performance, delivering generated music compositions and dance choreographs that are realistic and faithful to  ...  Recent research works have studied generative models for dance sequences conditioned on music. The dual task of composing music for given dances, however, has been largely overlooked.  ...  With the advent of deep learning methods, sequence-to-sequence models have been proposed for generating dance sequences from encoded music features.  ... 
arXiv:2201.11999v1 fatcat:ktgq7skannbkdiajjlm7rerb4y

Music-to-Dance Generation with Optimal Transport [article]

Shuang Wu, Shijian Lu, Li Cheng
2022 arXiv   pre-print
In this paper, we propose a Music-to-Dance with Optimal Transport Network (MDOT-Net) for learning to generate 3D dance choreographies from music.  ...  It has been tackled by different approaches such as similarity retrieval, sequence-to-sequence modeling and generative adversarial networks, but their generated dance sequences are often short of motion  ...  Another shortcoming is the inability to generate diverse dance sequences. [ Huang et al., 2021] proposes curriculum learning with L1 loss on the dance sequences to alleviate the motion freezing issue  ... 
arXiv:2112.01806v2 fatcat:dwguh76gmffhdj4fjfo4yz5stu

Hierarchical Style-based Networks for Motion Synthesis [article]

Jingwei Xu, Huazhe Xu, Bingbing Ni, Xiaokang Yang, Xiaolong Wang, Trevor Darrell
2020 arXiv   pre-print
Our proposed method learns to model the motion of human by decomposing a long-range generation task in a hierarchical manner.  ...  We first propose to explicitly disentangle the provided motion material into style and content counterparts via bi-linear transformation modelling, where diverse synthesis is achieved by free-form combination  ...  It is capable of generating motion with different styles (e.g., motion of novice ballet or modern dance of an expert) by optimizing with a cross-entropy framework.  ... 
arXiv:2008.10162v1 fatcat:yd4zkg64vzeqbf4xop7nzbfxv4

Music2Dance: DanceNet for Music-driven Dance Generation [article]

Wenlin Zhuang, Congyi Wang, Siyu Xia, Jinxiang Chai, Yangang Wang
2020 arXiv   pre-print
In this paper, we propose a novel autoregressive generative model, DanceNet, to take the style, rhythm and melody of music as the control signals to generate 3D dance motions with high realism and diversity  ...  It is challenging due to not only the requirement of realistic and complex human motions for dance, but more importantly, the synthesized motions should be consistent with the style, rhythm and melody  ...  Our goal is to generate realistic, diverse and music-consistent dance motion.  ... 
arXiv:2002.03761v2 fatcat:rkxfottbbrbutneemt4mwmedya

Towards 3D Dance Motion Synthesis and Control [article]

Wenlin Zhuang, Yangang Wang, Joseph Robinson, Congyi Wang, Ming Shao, Yun Fu, Siyu Xia
2020 arXiv   pre-print
In this paper, we propose a novel generative motion model based on temporal convolution and LSTM,TC-LSTM, to synthesize realistic and diverse dance motion.  ...  Unlike regular simple locomotion, it is challenging to synthesize artistic dance motions due to the irregularity, kinematic complexity and diversity.  ...  frames(30 frames), our model can generate realistic and diverse dance sequences with its own style.  ... 
arXiv:2006.05743v1 fatcat:nb2inkafqjfzlehap346q2zb7m

Dancing with cultural difference: Challenges, transformation and reflexivity in culturally pluralist dance education

Linda Ashley
2013 Dance Research Aotearoa  
Transformative learning and developing a reflexive view of dances from different cultures are pedagogical concepts that underpin the ANZC and I position them as key to exploring some transformations that  ...  My main aim in this article is to inform and support the development of an ethical and sustainable culturally pluralist pedagogy in which our responsibilities to the people whose dances we study are an  ...  different, and one would hope, transformational understandings of what dance 'is' for diverse people and cultures; • inform teachers as to how to integrate theory with dancing, thus avoiding a theory/practice  ... 
doi:10.15663/dra.v1i1.3 fatcat:5neqdubbczcojo2pz5c446gobm

Flow Guided Transformable Bottleneck Networks for Motion Retargeting [article]

Jian Ren, Menglei Chai, Oliver J. Woodford, Kyle Olszewski, Sergey Tulyakov
2021 arXiv   pre-print
Armed with both 3D object understanding and end-to-end learned rendering, this categorically novel representation delivers state-of-the-art image generation quality, as shown by our quantitative and qualitative  ...  Human motion retargeting aims to transfer the motion of one person in a "driving" video or set of images to another person.  ...  Our method can generate realistic images from diverse motion patterns. Youtube-Dancing dataset.  ... 
arXiv:2106.07771v1 fatcat:yjprgqyphbdenkfxguq5dpysxu
« Previous Showing results 1 — 15 out of 20,948 results