Abstract: This paper introduces V2Coder, a non-autoregressive vocoder based on hierarchical variational autoencoders (VAEs). The hierarchical VAE with hierarchically extended prior and approximate ...

Abstract: While transformers demonstrate outstanding performance across various audio tasks, their application to neural vocoders remains challenging. Neural vocoders require the generation of long ...

In this post, we'll trace the origins of the IBM logo design and its transformation over time. Follow the Inkbot Design blog for more brand histories. The iconic IBM logo is one of the most ...

We break down the 25 best negative space logos, from the FedEx arrow to the Toblerone bear. This isn't just a list of clever tricks; it's an analysis of what makes these logos powerful business assets ...

SAX J1747.0-2853 is an X-ray transient which exhibited X-ray outbursts yearly between 1998 and 2001, and most probably also in 1976. The outburst of 2000 was the longest and brightest. We have ...

RepCodec is a speech tokenization method for converting a speech waveform into a sequence of discrete semantic tokens. The main idea is to train a representation codec which learns a vector ...