AI Paper Atlas

Research map / Video / Video generation

TM2D: Bimodality Driven 3D Dance Generation via Music-Text Integration

Kehong Gong, Dongze Lian, Heng Chang, Chuan Guo, Zihang Jiang, Xinxin Zuo, Michael Bi Mi, Xinchao Wang

ICCV 2023 · Video › Video generation · approach: Text-to-motion diffusion · 92 citations (OpenAlex)

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add HumanML3D text-motion data to mixed VQ-VAE training (ablation)TM2D · Wild AudioFID_k: 43.85 -> 27.65Improved
add HumanML3D text-motion data to mixed VQ-VAE training (ablation)TM2D · AIST++FID_k: 23.94 -> 19.01Improved
add HumanML3D text-motion data to mixed training (ablation)TM2D · AIST++FID_g: 9.53 -> 20.09Hurt
add mixed-data training (ablation)TM2D · Wild AudioPFF: 0.78 -> 0.12Improved
sample tokens from top-k instead of argmax (ablation)TM2D · Wild AudioPFF: 3.76 -> 0.61 (top-k 1 -> 10, dance-only, wild)Improved
replace slerp transition with TM2D late fusion for text insertionmusic-text dance generation · AIST++MPD (ft=30): 0.135 -> 0.102Improved
joint music2dance and text2motion training in shared decoderTM2D text2motion · HumanML3DFID: TM2D (t2m) 1.021 vs TM2T 1.501; R precision 0.319 vs 0.424No clear effect
apply late fusion ratio 0.8 with music features at inference (ablation)TM2D text2motion · HumanML3DR Precision: 0.319 -> 0.300Hurt
use VQ-VAE shared codebook for both datasetsTM2D · AIST++ and HumanML3D tokensshared codebook vectors: 846 of 1024 vectors shared (98.9% AIST++, 92.8% HumanML3D)Improved
lower VQ-VAE downsample rate (ablation)motion VQ-VAE · AIST++ + HumanML3D (30% data)token usage and val lossImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy