Summary of Discord: Discrete Tokens to Continuous Motion Via Rectified Flow Decoding, by Jungbin Cho et al.
DisCoRD: Discrete Tokens to Continuous Motion via Rectified Flow Decoding
by Jungbin Cho, Junwan Kim, Jisoo Kim, Minseo Kim, Mingu Kang, Sungeun Hong, Tae-Hyun Oh, Youngjae Yu
First submitted to arxiv on: 29 Nov 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary DisCoRD, a novel generative model, bridges the gap between discrete and continuous representations of human motion. Traditional discrete quantization methods, like VQ-VAEs, struggle with expressiveness and noise artifacts, while continuous approaches face high-dimensional complexity challenges. DisCoRD decodes discrete tokens into continuous motion using rectified flow decoding and iterative refinement in the continuous space. This method enhances naturalness without sacrificing faithfulness to conditioning signals. Extensive evaluations demonstrate state-of-the-art performance on HumanML3D (FID 0.032) and KIT-ML (FID 0.169). DisCoRD is a robust solution for human motion generation, offering improved realism and efficiency. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about making machines that can create smooth and natural-looking movements of humans. Right now, there are two main approaches to do this: one is based on “bits” (discrete) and the other is more like a movie (continuous). Both have their problems – the discrete method isn’t very good at capturing details, while the continuous method gets confused by too many variables. To fix this, researchers created a new way called DisCoRD that takes these “bits” and turns them into smooth motion. This method works really well, making it a great solution for people who want to generate realistic human movements. |
Keywords
» Artificial intelligence » Generative model » Quantization
