English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

Deep generative models are often used for human motion prediction as they areable to model multi-modal data distributions and characterize diverse humanbehavior. While much care has been taken into designing and learning deepgenerative models, how to efficiently produce diverse samples from a deepgenerative model after it has been trained is still an under-explored problem.To obtain samples from a pretrained generative model, most existing generativehuman motion prediction methods draw a set of independent Gaussian latent codesand convert them to motion samples. Clearly, this random sampling strategy isnot guaranteed to produce diverse samples for two reasons: (1) The independentsampling cannot force the samples to be diverse; (2) The sampling is basedsolely on likelihood which may only produce samples that correspond to themajor modes of the data distribution. To address these problems, we propose anovel sampling method, Diversifying Latent Flows (DLow), to produce a diverseset of samples from a pretrained deep generative model. Unlike random(independent) sampling, the proposed DLow sampling method samples a singlerandom variable and then maps it with a set of learnable mapping functions to aset of correlated latent codes. The correlated latent codes are then decodedinto a set of correlated samples. During training, DLow uses adiversity-promoting prior over samples as an objective to optimize the latentmappings to improve sample diversity. The design of the prior is highlyflexible and can be customized to generate diverse motions with common features(e.g., similar leg motion but diverse upper-body motion). Our experimentsdemonstrate that DLow outperforms state-of-the-art baseline methods in terms ofsample diversity and accuracy. Our code is released on the project page:https://www.ye-yuan.com/dlow.

DLow: Diversifying Latent Flows for Diverse Human Motion Prediction