Welcome to #SaTML2024 ! In the opening remarks, we are excited to announce the two Distinguished Paper Awards!

First: Data Redaction from Conditional Generative Models by zhifeng kong kamalikac arxiv.org/abs/2305.11351 1/4

Nvidia presents Audio Flamingo

A Novel Audio Language Model with Few-Shot Learning and Dialogue Abilities

paper page: huggingface.co/papers/2402.01…

Augmenting large language models (LLMs) to understand audio -- including non-speech sounds and non-verbal speech -- is critically…

We release our code and model at:
github.com/NVIDIA/CleanUN…

A conditional point diffusion-refinement paradigm for 3D point cloud completion (ICLR’22)

TL;DR: SoTA point cloud completion method based on denoising diffusion probabilistic models and pointnet.

Paper: arxiv.org/pdf/2112.03530…

Project: github.com/ZhaoyangLyu/Po…

Speech Denoising in the Waveform Domain with Self-Attention
abs: arxiv.org/abs/2202.07790
project page: cleanunet.github.io

VocBench: A Neural Vocoder Benchmark for Speech Synthesis
abs: arxiv.org/abs/2112.03099

New blog post: Understanding Influence Functions for Interpreting VAEs

We look at what influence functions reveal about which training samples are responsible for high likelihood at a specific test sample.

Paper: arxiv.org/abs/2105.14203
Blog post: ucsdml.github.io/jekyll/update/…
1/n

On Fast Sampling of Diffusion Probabilistic Models
deepai.org/publication/on…
by Zhifeng Kong et al.
#ComputerScience #Learning

Neural vocoding just got faster with #DiffWave . zhifeng kong and @wpingnet describe a new, fast sampling method for DiffWave models without re-training! Code and pre-trained model are up on GitHub. github.com/lmnt-com/diffw…

Want to share our ICLR oral paper

DiffWave:
A versatile diffusion model for audio synthesis

It introduced a fast sampling method for diffusion model (trained with 200 diffusion step, sampling with only 6 step)

Project: diffwave-demo.github.io
Talk: iclr.cc/virtual/2021/p…

Patients prefer human doctors but recommend medical AI for others. Surprising.

Longoni, C., Bonezzi, A., & Morewedge, C. K. (2019). Resistance to medical artificial intelligence. Journal of Consumer Research, 46(4), 629-650.

Accepted as an oral presentation at #ICLR2021

📢📢📢 Our unconditional DiffWave can do zero-shot speech denoising now!! YES, unconditional generation of speech is useful as in language and vision! Check it out in Section V on our demo website: diffwave-demo.github.io